207
Divided attention effects on retrievril from episodic memory Myra Annette Fernandes A thesis submitted in conformity with the reqairements for the Degree of Doetor of Philosophy Graduate Department of Psychology University of Toronto 0 Copyright by Mym A. Fernandes 2001

Divided attention effects on retrievril from episodic memory

Embed Size (px)

Citation preview

Divided attention effects on retrievril

from episodic memory

Myra Annette Fernandes

A thesis submitted in conformity with the reqairements for the Degree of Doetor of Philosophy

Graduate Department of Psychology University of Toronto

0 Copyright by Mym A. Fernandes 2001

Nationai tibrary I*I ofcanada Bibbffi ue nationale 9 à du Cana

The author has granteci a non- L'autem a accordé une licence non exclusive iicencc d o h g the exclusive permettant à la National Liibrary of Canada to BibIiothèque nationale du Canada de reproduce, loan, distn'bute or sen reproduire, prêter, distri'buer ou copies of this thesis in microform, vendre des copies de cette thèse sous paper or electronic formats. la fome de micxofiche/film, de

reproduction sur papier ou sur format électronique.

The author .iretains ownership of the L'auteur conserve la propriété du copyxight in this thesis. Neither the droit d'autem qui protège cette thèse. thesis nor substantial extracts fiom it Ni la thèse ni des extraits subsîantiels may be printed or o t h d s e de celle-ci ne doivent être imprimes reproduced withouî the author's ou autrement reproduits sans son pezmissi011, am&&*

Divided attention effects on retrieval from episodic memory

Myri Annette Ferundes Department of PsycBdogy, University of Toronto

involveci in episodic memory rettieval, In each of the six acpaiments, participants studied

a list of random words under full attention, and r d e d than while paforming r

distraaing task presented visuaüy on a cornputer. Pmrious reseuch suggested tbat

divided attention @A) d u ~ g retrievai dirnrpts n# recall for words if the distracthg task

aiso imrolves memory for verbal mataial (Fernandes & Moscovitch, 2000). nie purpose

of the presait work was to detennine more precisdy which hctor(s) modulateci this e f f i

and influence r e t n d s u a e s . In the h t acpaiment I found that verbal distracting

mirs, that r@ed animacy or syllable decisions to words, produad large interference on

rnemory. ExpG-ts 2 and 3 showed that phoaemic decisions about nonsawe-words

produceb a &niMy large interference efficton mais-, that was $rgathan d c

decisions or size-estimations about piaurrs ofobjects. These findings support the

wmponent-proces modd of memory, which suggests that mtïievai W large dimpted

only when h r e is cornpetition for a common r e p d o n a l system. In the next set of

e q e r h ~ ~ en aitemative, reduced-resowœ, accuunt of Mtdkrence fiom DA at retneval

is considered by comparing the paf<pmaiux of yotmg md oid adults unda DA conditions.

odd-digit ta& did not produce as large an m Qtha grwp. F i , the contribution

ofthe h n t d and temporal lobes to interfefence effcctr on memory were examined.

Eldedy participants were dkided pmeqerhmtany Bis0 4 groupa, determined by th&

scons on munues of fiontal and temporal lobe ibndion, derivecl fkorn

neuropsydiogical testing. Large interfkrence e&cts on mawry were produced by the

aaimrcy, ôut kitt the odd-digit dhtmdng ta&. The pattem ancl magnitude ofinterference

e f f i did not dina depending on Id of temporal or fiontai b c t i o a These r d t s do

not support the hypothesis th& &écts of DA at retrievai are due to a reduction in gened

proceSgng resmrces, attentional capacity, or cornpetition for memory structures in the

temporal lobe.

AcLeowlsdgements

1 am vay g r a t a to Morris Moscovitch for bu advice aad encomgement

throughout my graduate school trahhg. 1 had such heipfbl and kind lab-mates and fiiends

while completin6 this work: a big th& you to Tonya, bhhne, Amy, M q m t , Sbayna

a n d m Iwu,hickytohpvebodtbetechnidpreistuiceofMirilynedElizabeth,

and 1 thaL than for dways being ready and wiiüng to heip. For teminding me of what is

important, 1 thank my parents. In particular, 1 un g r a t a to my mom for k ing there to

help take care of Martinique, anci giw w tirne to write this the&. Also, for helping me to

stay groudeci m reaüty, 1 thank my brotha and Leith.

A great big thank you goes to my M e h&arthique, who never tires of e x p l o ~ g

new~udplaces;Ihavelean#dalothmyou. I a r n e s p e c i d l y ~ t o M a r e ~

for ging me the idea of going to graduate school, for chrrllenghg m to do my best, and

for ahivays beUig thae (with a silly joke) to brighten my day. 1 couldn't have done it

without you.

Tabk of Contenta

Abstrrict

Acknowkdgemeatr

Tabk of Contents

Lbt o f TaMts

Lirt o f Flrires

List of Appendieu

Clmpter 1: Iaboduetion

The dd-task technique

Use of duai-tasks in other work

Review of literature on DA &eds at aicoding and retneval

A modd of memory retrieval

Ovemew ofatperiments

Chapter 2: Factors modol.ting DA t R i rt mtrkvai

ûvmriew of Chapter 2

W e n t 1

Expairnent2

-3

Ga#.l Discusgon of Chapter 2

Cbaptcr 3: DA i. yoiuy and old adda

ovaMew of chapter 3

-4

Expairnent 5

Expiment 6

Gaaasl Discussion of Chaptei: 3 Expaiments

Chaptcr I: Geneirl Discussion

suirmirry of Expaimants

Refemncu

Appendices

Tabk

Expefimerit 1 - Numba of Words Recaüed, Accuracy Rates and Reaction Ties Cui müüseconds) on Dishacting Tasks, and Percentage D e c h From Single to Dual-task Condition For Each Méasure

1 - Nwnber ofcorrect Respoases, R d o n Times Cui Milliseconds) for Correct Rwpoasas on the Conthow R e d o n Time Task for Each Condition

Experiment 2 - Numba of Words Recalled, Accuracy Rates and Reaction Times (ii müü*seoonds) on Distractiag Taskq and Percentage D e c h From S h g k to Dual-task Condition For Each Meastue

Expallnmt 2 - Number of Correct Rcsponses, W o n Thes Cm hdilliseconds) for Correct Respollses on the Continuou M o n The Task for Each Condition

Expariment3 -N~mbet~fWordsRecalled,A~~~~~Rates~~~d Reaction Tii (im miiliseconds) on Disti9cting Tuks, and Percentage Dedine From Single to Duai-task Condition For Each Measure

Eqmhent 3 - Nimber of Correct Responses, M o n Times Cm Millisecon&) for Comct Respomes on the Continuous Reaction Time Task for Each Condition

t<paiment 4 - Numba of Words Recalled, Acnaacy Rates and Reaction T i cm miîIiseconds1 on Dirrtrsdmg Tasks, rrsd Percentage Change From Siagle to Duol-task Condition For Each Measure for Young t a t s w=w Experbent 4 - Nimber ofwords Recellecl, Acniracy Rates and Readon T i cm mibconds) on D&mc@ Tsslrg rnd Percentage Change From Single to Duaî-tssL Coadition For Each Mersurr for older aduh (N=24)

Page

28

31

40

43

52

55

79

80

84

10 ~ t 5 - N u m b a o f W o r d s R c c p l k d . ~ R a t e s P a d 91 M o n T i (in dseconds) on DisLracting Tasks, and Percentage Change From Simgle to Duai-task Condition For Each Memm for older addts (N=16)

11 Mera Characteristics (M + SD) of Groups Selected Accordhg ta Frontal 101 Lobe (FL) FUlbCfion and Mediai Temporail Lobe (MTL) F d o n , anci Estimates ofPubüshed Normative Data on Ne~~op5ychoio~cai Tests for Aduhs Aged 65 to 75 Y-.

12 Eqerbnt 6: Numba of Worâs Recailed a d Pacentage Change From 105 Shgîe to Duai-ta& Condition For Each Mcuurr in Each Group

13 Ei<paimaat 6: Additional Words Recslled Following each Divided 107 Attention Condition for Each Group

14 Expiment 6: Acuuacy Rates on Distra;cting Tasks and Percentage Change From Single to Dual-task Condition For Each Measun m Each Group 109

15 Expehent 6: Reaction T i Çm miniseconds) on Dhachg Tasks and Pacentage Change From Single to Dual-tasic Condition ror Each Measue inEadi Group l IO

Meanpacenîagedecliaeinfiareullpedo~fiomfiülto divided attention (DA) for erch coQditr0on in Experiment (Expt.) 1,2, and 3.

Mean pacentage decîine in rcnincy rate h m single to divided attention (DA) for each condition in Experllaent mt.) 1.2, and 3

Mean pacaitage d&e in fke mail performance from fU to divided attention (DA) for each condition in Experiment m.) 4 and 5, in young and older aduits.

Mean percentage deciine in fi# d fiom sin& to divided attention for older amilts with reiativeiy high vasru, low fiontal lobe function with eachdistractingtpslr

Mean pemntage decline in aawacy rate f?om single to dividedattention @A) for each condition in Experiment (Expt.) 4 and 5, in young and older adults.

Mean percentage dedine in acair~cy rate 6om single to divideci-attention @A) in olda addts with relatively bigh versus low fkontal lobe fûnction for each disaacting tasic.

Expamient 1 Tnsüuctions to Participants and Expaimenter's tasks

2 büuctions to Participants and Exprhmter's tasks

Experhmt 3 3 d o n s to Participants and Experhmter's tasb

Expriment 43, and 6 Instructions to Participants and Expaimentds tasks

Instructions for auditory CRT task paformed concurrentiy with eachdistraaingtask

Sample data collection sheet for h e r e d

S ~ p k stimuli for distracthg taPlcs in Expairna 1

Smpk stimuli for distracthg tPPks in Ejrperiments 2 and 3

Sample stimuli br distracthg tasks in Expiments 4,s and 6

Page

164

171

178

184

190

192

193

194

195

=w-k

G e n d Introduction

T'hem are theoretid nasons for wanting to sbidy the effects of divided attention

(DA) on memory. The degne to which attentionai conditions iduence what we take in

and nmemba about the wodd provides a W o w on the capacaies and Limitations of

humui m f o d o n processing. A swprisiag ad count aiabiitive fiading in the rnemory

litamne is thrt e p i d c memory b disrupted cady w b attention is divided d h g

en&& but less so during retrieval. Demonstrations within a laboratory d g , ofa

debüitating eSect of DA on r e t r i d have been variable, ad sometllnes dificuit to

achieve. This is unexpecfed @en tbat most people allege that retrieving information fiom

memory, be it the neme of a movk, nuniliar fixe, or m e r to an exam question, is an

effortfui ta&, oftm thwarted by distraction,

The existence of an asymmetry m effebs, on encoding a d retrievai, is unacpeaed

on the buis of traditionai theones of memory. Smrsl infIuential theorists have suggested

that e n d h g d naievsl are similer or even identicai, ad they postdate a substantial

overiap between the two processesOceSSeS For example, Craik (1983) pro~osed that aicodmg

processes are d y those involve& in paceiving and comprehending, and thw r e t n d

processes oparte Ui reiostriting the ssme pattem of cognitive activity as at encoding,

thereby altowing recoilection of the eveiit, In a rgnilu veh, Tulvmg's d g spcaficÎty

principIe (1983; Tulving & Thomson, 1973). as well as the transfet appropriate prOCeSSiIIg

tbeory ('Monk, B d o r d , & Franks. 1977; Roediger, Wddon, & Chdis, 1989) also

aabody the notion that r e t r i d proasses must d e c t the specXc operations that were

CMied out miriag emadhg S i , Kolers' (1973) procedrirpüst view suggests that

W W e intemmoftheopas~ione .. cmKmbaLof-heaccthetwoorr

Wed. N n i r o - v g and Iesion studies aiso c o q e on the idea of simihnty in

d m g and retnevd proceslies: the same nanrl prthwiysysactivated when mfonaation is

perceivecf and taken in, an again readvated w&n Uiit idormation is n c a v d (MisMn

& Appenzelier, 1987; Moscovitch, Kapur, K O h k 7 & H d e , 1995; Nyberg et al., 1995;

Squire, Cohen & Nad4 1984).

There uq however, in di^^ tht some bnin regions ua Speaocslly active

duriag one process Md not the o k . For example, the lefk pdontal region has been

coasistently shown to be involved in eacodtig. whemas muiag retrievai, the right

praSontJ region is more hvolved (Nyberg, Cabeza, & Tuhring, 1996; Tulving et ai.,

1994; but see Wagner et ai., 1998b who founâ that for both encoding and retrievai,

infirior prefiontal activation was Iateralized bued on mitcrial type rather than mnemonic

opedon; and sa Raye, Johnson, Mitchdl, Noide, & D'Esposito, 2000 who suggest

retrievai may require interhemispheric interactions). Although there ir wodc showing that

the ri& pre-fiontai uea is involved in encuding ofnon-verbal materiah, such as

unhmniy fàces (KeUy et aL 1998; Wagner a al., 1998b). the mal-ority of studies indicate

tnit encoding a d retrievaîactiv8tions are seen in oppsite h m h e r e s . Likewise, if

éncoding and r e t r i d proceses are indeed simiIar7 mullpulatiotls that affect

one sat of processes shodd have a smiilrr &kt on the othr set, .Id not différent & i s

as some sf~dies have f d . B h h u n , Parka, Hartley7 and Nobk (1978) f d tht the

&~ofJcohalonmea#,ryeacodiagrreSteable,kitbeffea~onrrtrievllarrquite

Sm4 Afongtbe sameiines, Curraa(L991) showedthtbailodirzcp'mes have a more

negatk e&ct on encodiag than on &mai.

Wah respect to attentionai manipilations, it W wdl known thaî encoding is greatiy

affécted when attention is divided, leadhg to poor memory (Anderson, Cdc, & Naveh-

Benjamin, 1998; Baddeiey, Lewis, Eldridge, & Tliomson, 1984; Cr&, Govoni, Naveh-

Benjamin, & Anderson, 1996; Murdock, 1965; Naveh-Benjamin, C& Gueq & Dori,

1998). However, r dirent story emerges when the attentionai manipulation Y

introduced only at retned Some studies were able to demonstrate an interf'éreace &ect

with DA at rdncval, whüe otbas have f d little Euiy indidon of a ddetaious aect

(Anderson et pl., 1998; Bcddeley et ai., 1984; Cnik et al., 1996; Kdog, Cockh, &

Bourne, 1982; Naveh-Benjamin et al., 1998). Thus, the body of daîa to date demonstrates

that memory perfiomance is much more afféaed by DA at d g than DA at retrieval.

S t ü l w m d i f f a m a s h t 6 e & e a ~ o f D A r t ~ d r ~ m l w a e f o u n d

when performance on the distrading task h considerd S e v d -chers have 1~1ted

substantirlly greata costs to distracthg task pediocmatlce when it is performed

connnrently with re$deval (Johnston, Onenbetg. Fisher, & Martin, 1970; Tumbo &

Miione, 1971) u compared with encoding (Anderson et ai., 1998; Craik et pl., 1996;

GrifW, 1976; Johnston, Griffith, & W M 1972; Naveh-B.jpmin et ai., 1998).

Monova, as C d et al (1996) have shown, these rrynmetrid effkcts are mt a simple

case oftradeoflk between tasks.

hthis~IseeLto~efst8(rdtbedaiDptedSmagaad~le,&ec*rof

attentional manipulations on r e t r i d . 1 iIso p h to outhe the composent p r o a a w

mvdved in episodic w r y rehMnL Tbe &a&-task technique îs useâ to dow inferetlce

ofthtrepmceoseqby~theocamencedextentofmtaf~e~cee&a<on

r a r i e d , ~ m a c a ~ p a f o d t r d c Themigpabi&ofe&ctsfi.omvarious

conament ta&, t&at use differait types of materid d o r processes, are compared.

niese datr are used to deveiop a ~e~r~psychologicai modd of episodic memory. The

modd is then testeci by considering how ;igiag intacts with the observeci interference

&eda Further evaiuation of the mode& in tams of posriMe mediating bnin r Jons, is

carried out using a popdation of olda duits c M e d by kvel of dpninction in temporal

and bntal lobe sensitive tests.

Since aU of the stwües in this thesis, ancf the conclusions &am fiom than, rmke

use of the dual-task paradigm, the fint section reviews the technique and aunrmrim how

it bas been uscd in pmious work. The nad Secfion studies of the application of

the dual-tusk technique to the shidy of eacodllig and retrieval of episodic memory. The

fiaal section Oufjines the modd used to guide and interpret data in this thesis, and

descni the donaie for the series of e q d m s t s that fokw.

e du-k tecbniaup

The duai-tssk or M d e d attention technique can be d to aid our understanding

of cognitive fiinctoning by helping us to mfa the type of resources ad component

proasses d d e d by a pertiaikr tasic. Previous reserrcbas bave sucadùliy used the

tecimique to demonstrate that iftwo Cssb dnw on the same murces (Aiiport, Antoniq

& Reynolds, 1972; Bmoks, 1968; Famer, Beram, & Hetchei, 1986; Wb, Anderson,

Barker, Braüey, Fearneyhough, Henson, Hudson, & Bddeky, 19%), the same

hanispbae ('Friedman, Polson, Dafbe & Gsskg 1982; Klein, Moscovitcb, & Mgna,

1973; M-viîch, 1976; Moscovitch & K k h , 1980; Wicken~, 198û), or tdh ~ o t h e r

cornmon----ma L978;Khgbg&RoIand,

1997; nhrtin Wim Laionde, & bh& 1994; Moscovitch, 1994), interference wül be

o b d on one or both tsdrs whcn they ut p a f o d simuitaneously.

For example, s e v d studies have show thit diffhent distracting tasks produce a

distmCt pattem of interfiefence on targe (primary) tasks, b e l i d to be mediateci by

diffarsat brain regions. Mirtin et al. (1994) showcd thrt Ietter iad category fiuency, which

are mediafed primuily by the fiontal lad tempo4 lobes, rrspeaivdy, were di&rentiaily

âisrupted by distractin8 tasks ôeiiwd to reqyire the sime reswces and n d systems as

each of the fiuency tests. That is, an objea decision crrL mtafénd mon with category

than Inter fluency, a d similar to Moscovitch (l994), a sequential figez-tapping task

intdered more wïth letter than category fluency. These snidies suggest that it is possiMe

to iafar which brain regions are p r e f i iavofved on diffèrent tests based on the

pattern of interference SectS created ôy d h r a d q tasks. I f t k e is cornpetition for

common areas or resources, interfetetlce b c r d An examhion of the omouat and

pattern of interfaara obsaved unda duai-task conditions may indicate the degree of

owr@ in cognitive resources, components, anci structures @ed for the two tasks, and

a n be d to provide insight mto human maaory processiag

o f du&tasb in other work

The dd-task technique has been used extensive to sbuly the general procesSng

capaûty of humans. The singfeapocaY view of huinan information ptocessiDg (e.g.

Kdmmm, 1973; N o m & Boôrow, 1975,1976) was challeaged based on resuits from

aq>aimeats. W e some usb were sbown to be dif]Eicult to pafonn

aiiniltPaeously, such as shadowkg s g o b words and comprehending information m a

second bguage (Moray, 1%7,1%9), otha tasks codd k combiacd rrlPtivay e d y .

For e m q h , Mport et aL(1972) fwad M e kîerfiaena behucen shadowing of words

ChqMer 1: Iitrocsctioi 6

F- --

rad p l o . piano music, or fecognhbg pictures. Along the same lines. Brooh (1%7)

showed thaî feading, a visual ta& was more difficuit wben pafonned concurrently with

aaother task that nquired vitmakation, and Iess so with mother ta& for which no

Msdizatio~t wrs useci. Such results were Urompaa'Me with a sing1e-c- view, and

led researcbers (Auport et ai., 1972; Friedmrn a d, 1982; Kinsbourrïe & Hicks, 1978) to

propose tht huma^ have d 4 h a ~ e l processors .ad tbat it h oaly w b the spm

processors are involved simultaneoudy in two tasks that paforxnance wiU be &dt.

Kinsbourne and Hicks subsequentiy attempted to understand the resuits of dual-teek

acpaimaits by proposing that the degree of interference h m simuitaneous task is an

ifivene fimaion of the 9imdonal dism&' between cerebtal control centers. 1 wilî

nturn to this point later in the general disnusion as it bears on the mtaptetaîion of the

presait apaimeatil data on episodic mmory retrievai.

Tbe duai-ta& technique bas rlso ban used to d*amiae tbe characteristics of

short-term memory. S p d d y , the concept of 'working memory' proposed by BeddeIey

and Hitch (1974) &ew its support 60m various Qcperimmts showing seleaive

interfierence effects fiom duai-tasks- Their modd posas a limïtebap.aty centrai

ex- which coordinates the opedons of "slave" sub-systems, the phonologicai loop

d visuo-spatial skctchpad. These sub-systesns provide tempomy stonge for vabrl and

visuo-spahi mtezid nspectiveiy. Support for this idea came h m studies sbowing a

greata dWNgtBn &ect ofuriattended speech on short-term memory of words and digit-

span maddeiey, 1986) than on mernory of viriuo-qathd mformuion (Logie, 1986).

Comnrsdy, a visuo-spd tmckhg tiuL Ppoduced greater disniption on visual-

~based~thenontssksthatwerepudyverbr l@rooics , 1968). Fsrma,

Bamrn and Fletcher (1986) also f o d that verbal, but mt spathi, reasonhg was

considdiy impaireci by rote reheatsil of digits; in contrast, spath& but aot verbal

reasohg, was mon dunipted by a conairrent spatiai task. Thus, h is powibie to use the

duai-ta& technique to support the postdatecl existence of two Mèrent 'slave' sub-

systems in Baddeley and Hitch's (1974) mode1 of worlcing memory. Usuaiiy, intafernia

&écts from duai-tasks in attn'buted to compatiton for g d p r d g resources, but

as we see here, they csn a h be useâ to d e t d e the characteristics of aib-systans

imrolved in a model of worlting memory.

In this thesis, the dual-task technique is applied to the study of episodic manory

retrid. 1 use it to determine the component procesres involved in rrtneval, and to

develop a model of memory.

ew o f literature on DA ef f i~ ts at encdinn and retrievd

Reviow work used the duai-& technique O examine the murce requirements

for cnwâing a üst of items for a lata m a n g r test. Vdow distrading tasks such es card-

sorthg (Baciddey et al., 1984). digit-monitoring (Pu4 Smith, Dudley, & Laftontg, 1989;

Miirdock, 1965; Jacoby, 1991; Fernandes & Moacovitch, 2000). wordinonito~g

(FernarÈdes & Moscavitch, 2000). a visuo-spiitlll task (Cr& et al., 1996; Naveh-Benjamin

a aL, 1998), and fUc*, (Keiiog et al., 1982) hn been used. Despite procedurai

ciiffi- acms sbidies, each of tbese dkacting trsLs was capoMe of i n t e r f i with

encoding as indeKcd by the large dediae in memory pafommce!, nn&g b m about 30-

SV?, foiiowing DA conditions.

BccsusesuchaVanetyofdistnctiiigtsJawatapabIeofmterf~wah

mcoding C d et ai., (19%). Anderson d al., (1998). d NBVefL-Benjamin et ai. (1998)

conchidecl tbst s u d encodhgddeaisnds gened pmxsshg ~ ~ S O U ~ C ~ S . In m e n t

with t h spcailotion, Naveh-Beni and bis oolfeegues (1998) have recently shown that

the tadr d to divide attention during encodiag must utüize geaerai resoufces to be

effective at reducing memory. From a aeuropsychoIogical m e , Fernandes and

Moscovitch (2000) suggested that i n t e r f i h m DA at cncodllig arises h m

compcrition fw rrsources aeeded for mnsüous apprehedon of materid which is

necesmy for formation ofa manory trace. Any dkachg taok that diverts murces

necessary for conscious apprehawion ofthpt rnated, Win lead to poor enCodin& and

poor formation of a memory trace.

Studies using the dd-task technique at rdrrkd, bweva, have aot provideci aich

a consistent view of resource requinmentS. On some tests @YWM & Jacoby, 1990;

Moscovitch, 1994; Park et al., 1989), DA at retnCVBiI led to a substantid Litderence

on mmory pdomimcce (though not as m e as thot associaîed with DA at

d g ) . In otha shidies, however, DA at retrid had litde, ifany, &ect on memory

pafomirnce (Anderson et d, 1998; Badddey et al., 1984; Cr& et al., 19%; and Naveh-

Benjamin a ai., 1998). The lack of an &ect of DA ot &ed LP sufprising, &en that t

is expaienaxi as such an effortfiil pcess.

As the relativeiy smell or interference &èc& of DA at rrtrieval was

mmpc&d, 1 will examine these sbdies more d o d y . Baddeley et ai. (19û4) used the

chul-task technique to acPmme the role of atteinion in cev val of long-term epwodic

mmories. They d d e r e d the c&ct of DA on mmory tests such as free mail, paireci-

assmkte laraiog lad feco@on, wbik subjeas comumdy @ o d a ard-sorting

tu4 or s h m h m d y Md a digit-1OBd in mmd Thy conciuded rêtrieval was automatic

becaue, in d of th& apaiments. DA producd either no rhction or ody a stight

reduction m memory paformance, of about 5-10% less thui unda fidi attention

conditions.

Along the aime iines, Craüc a ai. (19%) haâ subjects paform nzt recaU, aied

d or recognition tests of memory simultaneoudy with a visrul ~ntinuow reaction

the tadc Only smaü interférence efikts on memory @ormance were obsaved,

regardiess ofthe memory test used; m e m r y costs were bighest for fiee r d (13%

d e c h f?om fuli attention), lowa for aieci-rd Md almost nii on a mgnition

test. Howmr, they also examiaed how DA affecteci @ormance on the d i a tisk,

astbismeuweplaorefîects~nsounesaeededforretrievai. Theyshowedthat

distracting taPk costs VEUied dependhg on the amount of enMronmental support o f f d by

the memory trslc COSU were patest fbr fhe (26% increase f?om baseüne), less for

cued recaii (14%), and 1- for recognition (?A).

To nirtha examine the role of attention in mediPtllig intafmnce &ects at

nirievai, Craik et ai. examined how performance, on the mernory and distrecting task,

varied depding on which tasL was ernphp9zed. Memory paformance (on each test of

memory) nmaiaed idatively stabk dropping only a d amount wmpared to nin

attention conditions, despite tdling abjects to pay mon, or Iess, attention to the

distnctiag tisL. Thu changes in allocation of attention haâ relatively liale e&ct on

retrievai. Paformame on the disüa&g however7 varied considerably dependmg on

hawmuchattentionwasallocatedtoit. Tbeseresuhsarems&rkcontrasttotbe&écts

they obraved whn mention ans dividecl .t ~. memory perfionnrace wKd

qaem&dy depeainng on the amount ofgttention aihaed to the disüadq rad

memorytask. niat~tbemorr~onwup~dtotbedistractingtosk,tbewone

niemory perfiommce became, and vice vasa

Thus in canbsst to encoding, which U seasitive to le* of avaüab1e atteation,

ntnevrl appears immune to disniption by DA Becawe there w m costs to the distracting

task. d these wae sensitive to attentionai manipilati~n, C d c et ai. condudesi that

retriend, desp'ne king immune to disniption, wur resowedemanding and did not

proceed automaticalfy. They suggest that g e n d attentional reSOUTces are mxmary for

plrcing the cognitive system into a rrtrievrl mode, and for vohmtrry strate& operations

thrt elaborate and augment aies provicied ciuring cetneval,

ûther wok by Naveh-Benjamin a ai. (1998) ctamined the d e n c y of memory

reSridtowiowtpskd~wbmpafodconaimntfywithavisuolCRTtaslc

They consideml whether DA at retrieval would afE& wmory pafomce, in a cued-

recaü puedigm, more whm Iow v m high kquency wor& were to be retrieved. Low

fkquency words are barder to r d (Gnge, 1976), and may rquire more search

prooar~es that oould be dhpted d e r DA conditions. Thy also wnsidered whetkr

ditFenia pe~ctptuai attnîirttes, usinga diffaent vasus same voice at retrievai as at study,

wouid iaaePPe the e&a ofDA Ngtha of these dpufations aitered the resilierrcy of

retrievai to DA conditions. As in previous stuclies, DA had little ifany Hect on retrievai,

qpdess of whether memory was fbr low or high ae<iuency words, or for items studied

in the same or dinerent voie as at shidy. There were, boweva, si@cant costs to the

~tpplSwhïchwieddepeadmgonthememorytaskdemaods. Thesestudies

strongiy wggest that memory r e t r i d U imnium to di9uptioa, ad nms obligatoriiy,

though it does consume g e n d attentionai nwuras as nadicated by costs to the

~ t a s k s *

modd of memorv retriev4

These rcsults can alsa be understood in the context of a neuropsychological modei,

which a s c r i i dinaait rnemory fiindons to differeaf bnin regions. The wmponent

that detemines whether interfixeme is cfeated fiom DA at retnevai, is the extent to which

retrieviil is dependent on strate@ procases mediateci by the PFC, and on associate aie-

dependent pmeses, m&ed by the medial temporal lobe/ hippocampus 0.

wharau tbape mediated ôy theMïLlH are modular, and require fewar g& resources

for its opantions. Examples of tnemory tasks tht arc fiontally-mediated include nall of

w o r i z e d lists (hAoscovitch, 1994; Park et ai., 1989; Stuss et al-, 1994). list

discximidon @ywui & Jacoby, 11990; Jacoby, 199 11, and reiease from proactive

inhiiition (Moscoitcch, 1989,1994). Under DA conditions, as long as the concurrent task

b itsdf- cud draws tesources away fimm the memory cask, interference

sâouid be observecl on the above tests. As pndicted by the d e i , sub- interference

efEécts h m DA a . retrieval wece fOuad uhg these tests; the &'kt U attniuted to a

reduction in generaî fesoufces a-ie tq and O& by the PFC.

d, 1984; C d et al., 1996; ad Na*Bmjamh et si., 1998). the memory test consisteci

of fkee nCan, ased recaiI or recognition ofa bst of unnLted words. Pafommœ on these

tests U o f h cli9upted by EnTuH &mage. but d y by fkontai demige (Miiners Petrides,

& Smith, 1985; Moscovitch, 198% Scbscter, 1987b). As mggesteci by Moscovitch

(1994)7 ift& fiontal lobe contdmtion to the memory tesi is m i . thar interference

effkcts from DA at retrieml wiîi be sxnaii, as retnevai crin be pediormed by the modular

MTM& wbich opentes relativeiy automiticrliy and obtigatorily. Consequently,

cornpetition for g d processitlg resources, dnwn on and orgmhd by the PFC, is not a

haor that shwld .&a memory @ormance on these tests. The ody rrsourca

d d g rspect of retrievsl on these tests Lies in estaûlisbing and rnaintaining retrievai

d e , as wdl as monitoring output; aich pmxsses are thought to be mediated by the

PFC, and under DA conditions, are reflected in costs to the distracting ta&.

In an effort to delineate furber the processes involvecl in retriwal of episodic

memodes, 1 UNes@ated in ptevious work wah my sumsor7 MomS Moscovitch,

whether there are conditions under which DA does interfie with retrieval on tests

mediated primarily by the MWH system (Fernandes & Moscovitch, 2000). Although

o h did not find very large interference efEiects on retrievai of a Iist of melateci words

(Andason et al., 1998; Baddeley et ai., 1984; Cr& a al., 1996; and Naveh-Benjemin et

al., 1998). close examination of the component-pt~ces8 modd suggests some conditio~~~

that m y lead to memory dunrption Rather than CO- for general resources, a

comament ta& d u h g retrieval compete with th structures needeâ to rrrctivete

the memory trace.

h r d i n g to th componart-process mdel, the actuai rrcovery of the memocy

a c n n s w b a i t b e t n c e m t e r a c t S w i i h t h e ~ m a p ~ d e d e c p b r y . Thetraceh

thou@ to CO& of an ensemble of newons m the neocortex thPt mediates the conscious

apaience during encoding, and forms the paceptual repreimtation respom'ble for the

content of the expience. This trace is b o d to the Mn/S fonning a collsciousness-

content packet t6at is the tecovered memory (MoscoMtc4 1995a,b). During rrtrievai, an

intanaliy-generated or exterminy-presented nie activates the bKIU3, wbich ads as a

pointer or index to the neocorticai neurons represemthg the content of the trace. A

concurrent tasic at retrievsi may dimpt eitba the EnlZRI W o n , or the neocorticai

representation. That ïs, if the two ta&s require acoess to the M U H system (naded to

nrçtivlte the neo-corticai representation), htcrférence may occur on one or both tasks.

Similady, ifthe taro tasks compete for the percepaial npresenîation that is part of the

mawry trace, intafefetlce moy also ocair.

At a funaionai levei, the cniciil dement in the first condition is t h the target

memory anci concurrent ta& jmth involve memow wherriis in the second condition the

crucial dement is the similarity in the of information that is processed in ôoth tasks,

regardlem of whether the concurrent task mvolves long term memory ( L w short term

memory (STM), or &ply perception. In my d e r work we tested these hypotheaes

(Feniandes & Moscovitch, 2000). 1 wiil b r i e mmmarh these experUnents as they form

the basis for the studies included in tbis P U , thesis.

In the nrst of the a<paiments, participants studied a set of 16 auditorily-presented

woids under fiül attention w k h they had to d firoeiy doud after a 55 second, fikd

deiay (15 seconds of arithmetic dEiilations ad 40 sec of a distnctiag ta&). htring

recaü Eo the DA condition, participants ais0 bad to monitor simuttaneously a üst of

viarally-presented words m a mmbg recognition test In oae DA condition, some words

waerepeatedafterataiagof3 aleûqdmthtothrDAconditiogttalagof7or

more, rmlring esch a test of short-term or long-term memory, nspectiveiy (see Tuhring &

CoIotl., 1970). We found tbat both DA conditions were successfùi in producing a

sub- decmnent in r d of about 3W in coqmison to the full attention condition.

That there was no différence baween the short and long hg conditions suggested that

MTL/H, which is imrolved in long-term memory but not short-tem memory, was not the

locus of the cornpetition &ect baween the target and dirtraçting memov tasks. Instead,

the reSuIts sugeesteâ that the locus of the &ect was at the levd of neocorticai

representationai systems involvexi in word paception and production.

This condusion was confimeci in absequent studies. Mead of usina ninning

recognition as the interference we used word-monitoring In an animPcy- monitoring

tadq participants monitored a list of visuPuy-presented words and i a d i d whever

words denotmg living ttiings occurred three t h e s in succession, The &ects of this

interfiring task, at acoding and retrieral, w a e compared with tbat of a visual r d o n

time tasic used by Craik et al (1996) (For detds, sa Fernandes & Moscovitch, 2000).

Recaii dropped by about 50.h whai digit-monitoring was p e d i o d concurrently with

en&& but oniy by about 1û% when it was paformed with retrieval. By con- the

.Ciimicy-monitoring task, wbich h d s b i h effects to digit-m~nito- ot d g kd to

a substadai 3W dmp in r e d w h pediorrned diiriag retrievai- Thus, the d t s

sboweâ that interference effècts fiom DA at reüievpl were material-specific, but the

effiécts at aLcodmg wae g e n d

Tbest fedu are coasisteat with the hypothesis nom the compoeeiit-p~ocess

mdei tbat iatanrmCe at retrieval cau ocatr when t h is competition for

e n a l stnicairrq presm&Iy in neoeom coqethïon for hdl"LA rtnichirrs

d o a not appear to pmduce interfetence. In CO- at mcoding, it W cornpetition for

germai cognitive resoufces thit iafgeiy detaiiines the c&as of DA on subsequent d;

nduction of conscious awamess by any meaas leads to poor formation of mernory traces

orailpamqthaebyacplpllüagthegenaPl&èctsofDAatQlcOdine. Atiarievai,

rccovery of memory traces via the M l U I U oôligaîory and automtic uniess a

comrmaitly-rttended stimuiw compstes Mth ropreseri tat io~~, ükdy in the posterior

neocortq tht e i k tom part of the recovaed memory trace or that are needed to gain

~ t o i f ,

The purpose of this theais is to detennïne mon precisely wbich varU1es an

respom%1e fbr produchg large i n t e r f i effects on retrieval. 1 test whether the Iarge

DA &èct at retrievai U materid or pfocess-speafic, and if- 1 attempt to detamine

which materials or process#r are impficaîed. In so doing, 1 hope to speafy the I o n i s of

interference effects at retried, and suggest possible underiying munl systems mediating

mmory retrievai. In Chspter 2,1 testeci wbetha cornpetition for vabpt rnnemonic

processes coatri'buttes to the pmdrtctionofbrge interfenaa enintsmidaVA atrctricvat-

I rlso consider w M e r it is the s d c or phonemic aspect of a distracting ta& that

iduences the size of the dect on ~I~RIIK)~~ tesneval. These data are used to deveiop

firrtéa the compom~t-process modei described above, by e g wheuia the neo-

corticai rrprrseirtptions, for memory of the word kit, are d c or p h o d c . In

Chrpta 3, I consider whether en alternativt murce uxormt ofDA &ects et rrrrievpl is

~ ~ b y a s n i i a i a g b w ~ ~ ~ t h e ~ o f f a n o f e n e c t S h m ~ ~

thitusediffaaitmateRalsmateRals Ifirrthaevahutethe~intermrofpo~~1cbIeb~regio11~

cbp$erI: I.tri8dllCfic81 16

medllting the &ects, by considering paforxxmce under DA conditions in a popuiotion of

olda dults, c W e d by lavd of temporal d b n t d lobe M o n .

In each of my a<paiments 1 used the mahodology and m o r y task employed by

Fernandes and Moscovitch (2000). The desttgn eiiminated the possibility ofmodaiity-

spcQfic int- since presentation of the material for the memoiy ta& at study was

auditory and recaii was vocai, whexeas the distradmg îasks were dways presented Wuany

via a cornputer monitor, and the respomes to it were d.

A d o u s fïnding in the mmory iiteraaire is tht episodic retReval, for a list of

UIllGiated words, W relativeiy immune to the adverse e&ds of DA manipulations. In

recent work, however7 I showed t h retrievai an be dismpted dependhg on the type of

materiai used in the coaaimnt task (Fernandes & Moscovitcb, 2000). These a<periments

ied to the suggestion thet retnevztt ocnns obügatody, such that recovery ofthe lnemory

trace U u d e a c d by concurrent tasks, accept those that compete for access to the same

neocorticai repte~entatioad system as the memory trace.

In this chapter I use the dual-task technique to d e this proposai in more

detail. Specifically, I wish to estabüsh what aspect ofthe matenpl in a distracting tasic is

respomoie for producing large disruptions in retrieval, 1s the key -or competition for a

common representation, or is cornpetition for MTLH wmory structures also needed to

produce interfereace? In an effort to fùrther specify the locus of interference at retrieval, 1

aiso examine whether it is the semantic or word-fom component of a distracting task that

modulates the amount of dimption tbat is o b d on rnemory. This informaîion is then

useû to provide Wght mto the resmrce quhmemsforepisodic memorgr*riewt.

Ovedew o f Chanter 2

In the first otperùnent 1 investigaîed whether it was necesssry to have a memory

component m the &tract@ ta& to produce intataence. Ai of the verbal d b t r a d q

taslo h m Fernandes and Moscovitch (2000) hd a memory which left o p the

~ ~ ~ ~ ~ ~ 6 ~ ~ o o u k l k d u c ~ c o m p c t i t i o n f ~ m c a r o r y

structures ntha then for a cornmon, word-based representational system as îhey

suggested. In this experiment, despite e h h a t h g the manoq component of the

distnallig ta& 1 found that on-üne procesSmg in wotd-based tcwks such as making

animacy or syUabIe judgeïnents, p e r f i i connimnuy with retneval, produceci iarge

intafaence &&S. Furthamore, these t h e S e i were not modulrted by the Ievel of

procsaring requind m the distnctiag ta&, wbher semantic or phonologid. niis

suggested that the word-fonn wpect ofitems in the distractjng ta& ratha than their

semantic content, accounted for the hrge i n t e r f m &kt on mernory.

Coosistent with this suggestion, Experiments 2 and 3 showed that even a

distracthg task invoIving nonsense-words, that have no d c content, produced

comparably luge effects on mmory. A semaatic distracthg task involving pictures on the

otha hnd, with no word-form component, produced a signifiant, but d e r , effect on

memory. This work aie0 showed that increqging the l d of diffidty of the picture-based

ta& h m Expriment 2 to Expairnent 3, did not in- the size of the interférence

sffbct. Thus, 1 conclude that retr id is afRected maxhdy when a distrachg trsk

rirw,tves pocessùis ~fwor~forms.

EXperim-t 1

Insroduction

ThU experiment ~ u g h t to detemine whether cornpetition for verbai mnemonic

proases contributes to the production of hrge htedereace effècts under DA at retrieval.

A p0ss1'ble confoundmg fâctor Ui Fanandes and Moscovitcb (2000) was th& 9 of th&

CL.p6#&PIctonrodPl&gDAe!fScctdntRtvrl 19

distrachg tuh féquired a rn-ry lad, which may a~cwnt for the large i a t d i

&kt tby obsaved on memory ( C e 2 0 ) . Thu the Me efkcts of DA at retrievai

potenthliy couid be due to cornpetition for memory structures rather than for a cornmon

representationd system. Recognition ofwords previoudy seen in a long list ofvisuaUy

presented items wu one of the distracthg tasks used in thar study (see Expriment 1

Feniandes & Moscovitch, 2000). Such a task requires a signincant aemory Ioacl, as

participants have to keep several items in rmad while aicoding new items. This issue was

addressed by showing that a word-monito~g teslg which reduces the manory load for

the distrachg tadq dl cnated large interference e f f k c ~ at retrievd, whiie an d o g o u s

digit-monitoring ta& produced rdativeiy d e r e&cts on memory. However, it is stül

possible that Merence amse due to the mawcy load demands of their distractiag ta&

rcither tbui the vabd component, 40 Fernandes ad Momvitch (2000) suggested. It may

be t h keeping even a couple of items in mind for the word-monitoring trsk introduced ri

memory l d wbich, perhaps in conjmction with campetition for vabal paceptual

repmenîations, led to the large effécts ofDA rit ntrieval. By eliminating the memory

load compoemt of the ta& it shouid be possiible to determine whether or not it was

aucial for produchg large disniptions in r e t r i d pafomance.

The obiestive ofthe prese~t expriment wa9 twofold. First, 1 tested whaber a

disaading task thst involveci wordq but no memory 1004 wouid di p d u c e large

mtafamce on memory paformance, 1 acited two d i f f i wrbd distriictiog tasks for

&s purpose. One was an mbacy deQsion tasic in d c h participants decided wh*ha

each word in a li8t denoteci a liviog or a men-made object. The other was a syiIab1e

d c & k m * i h w h i a h ~ ~ ~ ~ & i r t a k k d r a i o q m ; i b l c r

If materiai-spcaficity is the CIU& &or producing large intesfk- &ecf~ ofDA at

retrievai, duchg the memoq ioad of the chacthg ta& should not duce the sïze of

the i n t w ef&ct. Such a finding would also support the c h h that intaference at

r e t r i d arises fiom cornpetition for a common repte~entationai system (Fernandes &

Moscovitch, 2000). I$ howeva, e h b i n g the t h e r y load eliminrites or reduces the

Uaafae~ce efFéct, them I wül have shown tbat compczitiolt for memory systems underlies

the &kt.

1 also wished to investiete whst aspect of the dLPtnctmg task modulates

interkence effects on memory. 1 considered the type of word processing tequired in the

di- ta& as a potential factor infiuenciq the DA effect, I compareci the animacy

and syllablc distracthg tasb os they required semantic, and phonoiogical processing,

nsp&wdy. Ifit is the semantic aspect of words th& modulates the intafaence &kt,

thm a h a q decisions in the distracthg ta& shodd leaâ to a télativdy greater &kt than

phonoiogical decisions in the syllable distndmg task.

In cddition to the effects of DA on manory performance, pmious studies ais0

considaed pafofmance on the distnctrsg tadi Cdk a at (1 996) and Iohnston et at

(1970) suggested that the tetrieval process daneiads signiscant general processing

resources, as indexai by the lPge dectenrents in d k a d q ta& @ormance unda d d -

tssL conditions iit retrievai. Contrary to BadcWey et A's (1984) daim that retrieval

ocnin ~utomnticegr, Craik a d. (1996) ress~aed tbst mmory retrievat ocam

obligrtorify (h they showed ody d eSxts of D A on memory), yet demands

considarbk resources. Whüe Fanandes anci Moscovitch (2000) showed that retrid is

not obügatory, a least whm the matai.ls or pmcessas used m the manory and distracthg

task are amüir, thy conw with Craik et al. (1996) that rrtMvll W resource-demanding8

Fernandes and Moacovitch (2000) suggested tht establisbing and maintainhg retrieval

mode, d o r monitoring output (not the asm&tiv~edepaident process or eqhory) is

the g d resource-demanding PPpect of d e v a i (anci tht cos& to the disüxdng ta&

are ïncurred regardless of ollailuity in material m e e n the memory and distracting ta&.

The p r e ~ n t a<paiment a<aniimd whether ncall, which is determineci more by semantic

thrn phonanic appects, would have a greata &ect on acairyr or M o n time (RT) in

the semantic than in the phondc distmdng ta&

Because the distractin8 tada themdves are tesoutcedemanding the amount of

generai resowces each uses may contniie to the size of the interfience &éct8 To

~isses, the dative resource dawids of each task, 1 looked at the effkct eadi had on a

concumntly paformed auditory coatinuous &on time (CRT) ta& In the CRT tesk

participants hd to iden- cornputer-generated tones as snher low, medium or high

pitched tones. The RT aiad n m k of comct respoases in the auditory CRT task were

r~rdeddMalyzed~amepiisof&au~howdenisndingerdidistractingtask~

with longer RTs indicating greater reswrca demanch.

Method

Partici-

Participants were ssked to try to commit un auditorily presented iist of words to

memory ad aibsequentiy a fkee d ta& was administaed as the target g e t r y task

Raor to rebitevd they began a dlsn9cting ta&, either aamiacy or sylfable decisions about

words presented vhaüy on a cornputer screen. in esch of the two DA conditions,

participants continued to paform one of the distracthg tasks wbüe simultaneously trying

to d out loud the target task word ti* Participants rlso pdomied a baseiine (fU

attention) condition, in which the distncting tasic aded prior to net r d .

Mat#irJs

fatpet recall Stimuli for the target rnemory tasks consisted of 64 unreiated

cornmon wuaa Words for the target manory tasic wae recorded in a s o d - p r w f booth

onto an d o file via a MacIntosh cornputer ushg the Souad Designer II program. Four

word IWu were created by randorniy choosing 16 words for each üst from the pool of 64

worâs, Each word iist was created with 3 seconds of silence mserted between words.

Three beeps wae also ncorded pnor to the beeiamng, and at the end of each word list

The lists were then recordecl oato an audio-tape end presmted via a cassetîeplayer. AU

n tasks

stong, Four 50-word IWts, ad one 20-word üst, wnsisting of words

r e p r d g animais and man-de objeçts were created fiom a pool of220 words. Each

list was aeated aich that M o f the words represented raimals Md &&man-made

abjects.

SvUabie decisions. Four 50-word lists, ad one 20-word IUf consishg of one,

two and thra syiiabIe words were dso created fiom another pool of 220 words. Each üst

was created such that haif of the words hd two syuabfes, and the rest had one or thtee

syliables. None of the words âom the syUab1e and riiimscy task were used for both t h .

Also, none ofthe words in the syilable ta& repre~ented animais.

AU stimuü were medium to high fkquency words chosen fiom Francis and Kucera (1982).

Word fkquetlcies ranged h m 20-100 occunences per minion.

Words for the animacy and syhble desision taslr were presented v i d y on a

cornputer acreen at a rate of one word mry 2 seconds. For the Mimaçy participants

Wcatcd ïfpresented words represenieda manma&e objest, and for the fiable bey

iadiated ifthe p r d word had two sylIablas. by pressing a key wah the dominant

wurithg band.

AIthough 1 recordai menual respoase times in ail of these errpaimeats, 1 did not

emphke to participants îhe importance of tespombg puSckly on the distracting tasks,

w b a i p a f o d sin& and indwl-tsskcoQditionswithrrtrierrL NeverthelessI adyzed

these data as they might provide an iiadirrct iadicator of kvd of diflEiaiity or processing

d d s of each dismahg task.

c task. The shidy phase for the target rnemory task was always followed

immcdlltdy by an aritbmtic task to elhimue ncency (as in Craik et d., 19%);

Participants hard a two- or threedigit mimba at the end ofeach word list, and were

instruded to count a h d backwards by threes. The digits were recordeci ont0 the audio-

tape in the same mamer as the words for the target maaory task.

Procedura

session. Participants were tested individuaüy. They pedionned the target

memory ta& unda ni1l attention, foliowed by a practice session of the anMacy and then

syUob1e distracting tasks. The study phase for the target taAc was identical in the practice

aid phase except bat dini t iists were studied for each For the memory

ta& participants heard a tape-ncorded W e voice reading a üst of 16 words at a rate of

approxhately 1 word wery 4 seconds, and were uked to try to commit the words to

memory for a lata recail test Participants then couated backwards by thtees stamng with

the digit spoken at the end of the word list for 15 seconds. In the ptIlCtice phase, r d of

the studied words ocaurrd immadistely foliowiag the anthmetic ta& Participants had 60

seconds for fiet r d . Participants wae then given a practice bIock for the aaimacy and

Uren @able dkractor tasks.

sessiom. S-e-task pafornrrace for eitba the animacy or syiiable

diemctnig t.sk was measund before any of the arperimental conditio~. Single-taslc

p a f 9 r m s a c e f o r t h e ~ g ~ t s s L w a s m e a g u n d at theend ofthe final

~terbalancedmssputicipaats.

Following the first shgie-task xneasure, the thne experhmtai conditions (full

attedon plus two DA conditions) were administered, co~nferbalanced ~CIOSS participants.

Presentation of the words for the target r d Eask was foflowed by the arithmetic ta&

tbaieitbatheanimacyorayUabIedistractingtaskbegan. Thedistractiagtaskwas

pediormcd alone for 40 seconds un@ the c o m p t a &ed a low-pitched torse. The tom

signaied that r d of taped wordr should b e g h For the two DA conditions, the animacy

or syllable task conthed on the cornputer whüe participants simultaneously tried to r d

words for the target taslc The distracting and net recall tasks were pdonned

simuitaneousiy for 60 seconds, and participants were told to divide their efforts e q d y

between the two tasks. The importance ofplecing 500/i oftheu &ort on the r d task

and 500/. on the distracting task was emphrsited Atta 4 in the DA conditions, the

expaimenta asked participants if* d e d any additional words f?om the study üst,

now that they did not have to do two things simuitaneowiywiY Participants' r e d responses

were trporemrded.

h uie 'fûît attention' e x p e r h d c o ~ t î o n , the distractiag task end& after the

cornputer sigded tbat fine recall h u i d begin Thus. h this condition recrll ocaimd

underni11sttentioi1, rindservedluabesdinewithwhichtocompue~rdintheDA

conditions. It shouid be mted thar in this condition Qtha the anhacy or syllabk tasic uns

p a f o d as a 'iüier' for the nrst 40 seamds (us& the 2thword kt) &a the mdy

phase ofthe target word list In this way the time hg between wbea the words for the

dtiuLwanstudied, aswdlasthemedtopaf~m~~i~)thertsdrbaorerrall, wen

the same as h the DA conditions. The 'filier' t.dr d e d once the computer signaicd that

recall of the taped words shouid kgin For ail ordas of #paimaital conditions,

participants were givm a four-minute break More begeming the next condition

f the distractinn tada. Ç o r n h a rtsource demanâs O

nie auditory CXT tasL was used to compare the murce derrmds of the Pntnaçy

and syllrible distracting tasks. For this taslc, participants had to id- cornputer-

gaiaated tones as either low, medium or hi@ pitched tones. The tones wae played in a

d o m orda, and patticipants were told to press the appropriate key as quickly and as

acctmîdy as posaile to identify the tom on each M. A new tone was pnsaited as soon

as the participant pressed a key, or &a 3 seconds had drpsed.

Each puticipoat wmpleted thne sessions of tbis CRT ta& as the final phase of the

arpaimait. The ta& was performed done for a bwline rneaswe, and in DA conditions

with the anhacy and @able tasks. For the DA conditions, ia order to avoid having

participants d e Mirent manuiil rrsponses for the CRT and distnctiag tasks,

participants rnade a verbal nsponse to i d e words (-made words for the anllnacy

task and two-syiiable words for the syiIable ta&). The #perimenter recorded the

priiticipt's ltsponses by pressing a key on a separate keyboard.

In the DA conditions, the tone ta& was @tmed alom for a short time, &er

whichoneafthdishsaiagtasksbcgandW100~ec0iads. TheRTendnumberof

correct res~onses in the auditory CRT were record& and adyzed as a m e ~ s of gapging

RÉsuits

Tmzet maaocy tasic. Both the anhacy and Syuoble distrPcting tasks mterfèred

substuitùlly with h e recali @ormance. There was no difference in the mgnitude of

this encct depending on whatha the distrsctiDg ta& was iaimacy or syllable decisions to

worda ThemeuisforeechconditioneeprrsmtdbTtble 1. ThedatiweredyLedin

a two between (order of acperimental condition and orda of singie-task measure for the

animacy and syIIable task) and one within ( e e n t a l condition) ANOVA There were

no signifiant main effécts or interactions with the order &ors, on fiet r d

pafocmance. The foilowhg d t s wae Sgnincant at p < .O01 unless o h & noted.

was I main e f f i of apaimental wndihon E (246) = 21.70, = 2.22.

The mean numbefS ofwords recaiied in both the animacy and syUable DA conditions were

reduced sisnificantiy fkom the mean in the fidi attention baseline condition, E (1,23) =

56.62, MSE =3.01 andE(1,23)=31.l0, MSE = 3.62, respedveiy. The diffierencem

numba of words d e d in the laimPcy versus syiiabIe c o ~ o n s was not significant, E

(1,23) = 0.90, a = 6.70. The mepn pemenhge dèche ih wor& recaiîed fbr 0 t h

participant, fiom full attention to DA cocditions, did aot differ for the two DA conditions.

Eqmhent 1 - Numba of Words -4 Acairacy Rates Md Reaction T i (in miîhcoads) on Dhtmchg Tub, and Penrentage Deche From Singfe to Dual-task Condition For Each Measure

Measure and condition M SD

Tu@ memory ta& Words tecatled

Full Attention 7.46 DA MimPcy 4.79 DA syllabk 5.29

P-e m m in d o n thne DA uiimacy 16-94 13 -60 DA syUabIe 1.74 19.12

Note= DA = dnnded attention

c h p t e ? ~ F I E t O C I ~ g D A c n c c t I f ~

Foliowing each DA condition, the partiaprnts were &en the chance to d

wo& h m the mget ta& unda fidi attention. Few participants d e d any additional

words. The mean auabas of additional words d e d &a the 'aaimacy' and 'syiiable'

DA conditions wexe o d y 0.54 m.= 0.83) and 0.54 (%DL= 0.93) respedively.

4sdamUm-

Acairacy rates (cainiloted as #hitsll5 - # fiilse alarms/l5) on the Dmmrcy and

@able distracthg ta& in the DA condia'ons, were much worse than unda single-task

conditions. nie percatage deçline in accuracy rate, fkom single to dual-task conditions,

was hger on the syllable than animacy taslr but the differetice was not significant. nie

data were anaipd in a two between (order of acpamiental condition and order of single-

tasic masure for the uiMicy or syii8bIe task) a d one within (a<perimntai condition)

ANOVk Thae were no si@umt main effkcts or interactions with the order hi*ors on

distractin8 tasic pafofmance. There was a main &kt of experimentai condition, F (3,69)

= 19.45, MSE = 0.02. The mean ~cairacy rates for each tasic, in each condition, are

presented in Table 1.

The mean urziracy rates on both trsks under DA coaditioas Mered Sgnificantiy

h m theh nspcctivc singie-ta& brseline wonnonce (E (1,23) = 30.96, MSE = 0.03 and

E (1.23) = 27.12, MSE = -05, for the i<imiacy and syüable tasic respedvely). Paccntege

deciine m curwscy nte, fiom Sagle to âuai task paformance, did not differ significantiy

~~~ssdwtractmg*.

Tirmet

significantiy lwga fbr the syilable than animacy distnallig tasic (r (22) = -5.49). The

hcmase in RT, fiom badine to duai-tiulr conditions wu, Sgnificantly greater

for the aaimrcy tban syüable task (g (22) = 2.99, p< .QI).

The cort.elation between RT on the di- ta& and memory interférence

Pnda each DA condition was non-signifiant.

m T

Distm&g&& The ciccurscy rate for both of the distrachg tasks d é r e d to a

stoiilar degree whai the auditory CRT trslc was pediowed concurrently. There was no

e&a of ta& order on acamcy rates. The mem accuracy r ~ s on the animacy and

syl1abIe tislrq paformed concutrentiy with the CRT tone task, were .44 ( ' . = -23) and

.46 @.DL= 22) rrspectiVely. The diffience between these two acairpsr rates was not

s i ~ c a n t .

CRT tone task. The diffaeace in the number of tones c o d y identifid in the

rnimPcy and syllable DA conditions was not sipifitant. The mean number of correct

respoases for eich coaditr'on are preseritedin Tabk 2. A witnui m~pant ANOVA

meaied a mUn e&ct of condition E (2,40) = 82-17. M a = 82.08. Thae were no main

eEects or interactions with the ords &or. PPkaaed cornparisons showed the number of

tones comctly identifieci in both the animacy anci syllobk DA coadaions differed

sipifiamtiy f?om the ni1l attention baseürse condition, F (1,20) = 169.17, MSE = 102.69

Expsriment 1 - Nimber of Correct Responses, Reaction Times (in MiUiseconds) for Correct Responses on the Coatinuous Reaction Time Task for Each Condition

Correct response

Condition M SD

Reaction time

M SD

Note. DA = divided attention

The mean RT to iden* tones is shown for correct responaes ody (see Tabk 2). An

outlier dysis eIuninated RTs greater or lesser than two standard deviations fiwn the

mean for tach participant in each condition. A within participant M A reveded a

main &ect of condition F (2,40) = 28.64, MSE = 9537.1 1. There were no signincant

main e&kcts or i n t d o n s with the orcier fàctor-

The mean RT in both the rnimrcy and syuable DA condition differed siBnific8ntly

âom the mean in the basebe condition, F (1.20) = 58.3 1, MSE = 128 19.02 aad E (1,20)

= 34.77, MSE = 25466.43, rrspeaivdy. The cüfRiereace in RT between the ammicy and

syUnbte DA condition was aot signifiant, E (1.20) = 0.3 1, suggesting the two

tasks d e simi7nt resource demands.

E v a n t h o u g h t b e w m o r y t o a d c o m p o a t m o f t b e d i ~ t r p k ~ ~ ~

~einterferencee&dswerestülobservedunderDAatretrid niemagnitudeof

these e&cte war 8miliu to those reported by Fanuvles and Momvitch (2ûûû), whae a

vabal nmning recognition task d a word-monitoring dhacthg task led to an

appmxhate 3û% deciine in m a w r y pafonnance, h m fuii to DA conditions at tetrieval.

These effécts are stibstanticilly k p r than those nported by Baddeiey a aL (1984), C d

et d. (1996), Naveh-Benjamin et aL (1998), and Anderson et al. (1998), but they used

distracthg tash that involvd non-vabai materialserials These d t s suggest thPt word

rrtridduringrrcalluuibeim~dsisnificentiybyaconairrenttriskthntalsouses

verbai material, Uraspoctive of its memoty Id. htdetence does not arise Eorn

cornpetition for memory-Speanc systems.

These findings are consistent with the hypothesis derived nom our

neuropsychoIogid d e i , that succcsaail retrieval repuires d g a verbai

representationai system and that DA effêcts at mieval when such a system is

eimultaneousiy requirsd for another task (Fernandes & Moscovitch, 2000). Becsuse the

semantrCc snd phonemtic concurrent tasks prodûceb e q u i v a ' interfèrence effects, I

conchde tbrt cornpethion may ocnir at a pre-semantic ieveî, pahaps for phonmiic or

word-fbrm repmentatiom. This hypothesis is evahmted fiirtba in -ts 2 and 3.

IfoundtbatpaformanceontheaciimscyrndsyUabledisaactingtaslrswar

sigdbdy affccted in the duai-tadE waditions. The magnitude of interférence on the

~rndsyUabiedistractUigtasks~consistentwahthatob~edhstudiesthat

d n o w a k C é i s a s t i i i g t r s l r q - ~ I M ~ - o a ~

p a f o ~ . As originalhl suggested by Cnik et ul. (1996) and Johnston a ai. (1970),

t h e s t n s u l t p s h o w t h o t t h e n t r * d p ~ d ~ s u b ~ g e n a a i p r O c e S S i n g

resources, as indexed by the poom pdorniaace on both dirrcractmg tasks under DA

conditions, ad daes nat operate automafiaEty as Baddeiey a ai. (1984) ciaimeci. These

&ts are a h in üae with the suggestion tbot establisbing and mpintaining rétnevai mode.

andmonito~output, m i y b e t h e ~ ~ ~ 9 p e ç t o f r c t r i d ( F a a s n d e s &

Moscovitch, 2000).

A direct anaiysis of task diflbiculty ( i i n d e n t of paforrnance under DA with

retrieMi) showed that both the aniniacy and syiiable distracthg tpsks had similar C n i on

the number of tones idemtifid and RT on the auditory CRT task* Moreover the acauacy

rates in dd-task conditions with the auditory CRT ta& were sUnilar. This suggests thPt

the two dhtmthg tasks were quaiiy difficult and resource demanding.

It shouid be aoted that Malysis of b h e RT to respond to each distracthg task

showed longer RTs for the syllabie ta& and that the percent inmase in RT, fkotn single

to DA conditi011~ with rrtnevai, was iarger for the mimacy ta&. Thus there is also a

suggestiCo~ ofa task-specXc wst in th t KT on t6e iniirYCy ta& was disproportionatdy

affebed by DA Because nspanse speed wu, wt aaphrisiad in the inseuctions, RT to

respond on the distrachg ta& may not be as sauitnre a m e m m as the auditory CRT

ta& whichIchoseatplicatytocomprirrnsource~eas~fortbedistractiig~

(as in Fernandes & Moscovitch, 2000). These r d are consistent with the c i a h that

~ t u L c o ~ u n d a D A w Q d i t i ~ ~ l ~ w i t h r e t n * ~ ~ i a a n n d ~ e s s o f

m a a t a t & m * d i s t n a i a g * m & m 2800h €hwgi?t8errb

miy be a process-specisc component a f f i the size ofthe dismahg task interference

efféct.

Ovarll, the uiimrcy d @able dbtm&ng tula produad similerfy large

intafaaice effects on memory, d e r c d to a similor degree wbai paforrned in duai

conditions with retnemi, and did not M m in their reaource dernands. Because the

fe8owce deman& of the distnaiag todrs were m e n t , a possiile herpretation of the

nnding is that retnatal intafaence depasds on the tesource daaends of the distractiag

tasic. 1 do not endocse this view, and instead favour a material-specific over a resource-

cornpetition rccount of &d interference. Even though the digit- and word-

monitoring tadcs used in my pmious work (Fansades & Moscovitch, 2000) were more

demaading than either ofthe tasks used in Experiment 1, intafirence &ectS comparable

m sut to those in this acpaimait were obtauied, ody in the word-monitoring DA

condition, supporthg a material-specific interpdon. This issue is explored fùrther in

subsequent experiments.

Experimrnt 2

Introduction

The cornmon -or across di acpaimeats in which large interference effects h m

DA at reüievrl werc found Cui Fernades & Moscavitch, 2000. as wd Expzhnt 1 of

thepresenttheoir),isthatthittlaedishPctingtaskimroLvedvabilmataùl Iathenexî

expaimeat, 1 investigated what aspect ofthe words m the distncriag ta& is reqoll~~cble

t b s ~ t k c ~ y ~ e f F c e t L L u L - k k t h ~ p r & o f k

phomiogicd procashg of word-&rm% that undalies b r f k m m ot retrievat,

To aâhss thU question 1 comprred the ïnmférence &ka on memory prduced

&om two different DA conditions a retrieval. In one, th disEractmg task retained a

semantic component, but had no word-&cm, and m the other, tbe distracthg task retained

a word-form, that requùed phoaological pmashg, but hd no semantic component.

Oat ofthe diûmchg (Isks r e q g h d -pan& to ideatify M w i r i g s npresenting

man-made objects, among line drawings of üviag and man-made items, whereas the other

requlled them to identify 2-syilsble nonsense-words, unoag l-, 2% and 3-syUabIe

pfonounceable nonsense-words.

Recaii tbrt ExpWnait 1 showed the levef of prooessing rquired in the dismaing

task did mt influemus the size of the Uiterfèrence &a. This leci me to suggest that

cornpetition at a pre-semantic level, paheps due to p d g of word-fonns, W

tespomile for the large interférence effècts on wmory. If& proposai h correct, then a

distrading ta& that involves phonologid procesSul8 of word-ke material, without

semantics, shodd stül impair retrieval. Along the sam lhes, a distractllg Csslr that

c011sl~sts of pidure mai& (no duea word-for@, saripntic pmcesSmg

s h d d produce much l e s interfixerice with retneval,

As in Expriment 1,1 also considered changes in rr;auocy rate on each dishaaing

ta&, pedorrned siasly and conamedy with retrievai- I a<pected sig,cant declines in

@ormance d e r dual-task conditions since r e g i d U a r e s o ~ e m a n d b g pmass

(Craik et al., 1996; Jolmston et al., 197Q in tbet rrsources are weded to establis& and

niriatriiirrtticwaieods&igObitoc~u$ut@--*2(##Lr

FurthermOrc, because the nonsense-wod rad pi- trsks Unrohnd diffefent materiais, 1

dso tateci the clah tbat distraaiiig usL costs are Uicumd, uDda DA conditions with

d e v a l , ngMUess of maten9al used in the dbtmtbg tssL (Fernandes & Moscovitch,

2000).

Because 1 was examiaing wbether the picturaanmiacy end nonsense-word

d k w t k g tasks ptoduced 9raütr zunmnts of iaterfuenœon mmiocy, I wanted to msure

that any diffaences wae not due to d i f f i resource de&. As in Experiment 1.1

d resource demands by comparing paformance on each distracthg ta& p e i f o d

concurrently with the auditory CRT tasic. Again, the RT and number of correct responses

on the auditory CRT ta& were remrded and anaipd as a means of gauging how

detnandjq ePch distncting tasL wkq with 10- RTs indiPafine a greater resource

requinmnt. For consistesicy with Experimcat 1.1 also analyzed RT to respond for each

distrabirg ta& paformed siagly and in duai-task conditions with retrievai.

Method

Particban&

Padcicpants w a e 24' nahe undiergndibate skrdenfs at the University of Toronto

who rrcQved $10.00 for th& participation. An pertiupmts claimed to be native En*

speakers, and to have normal or corresteci to normal vision and hearing.

.. decisions. Stllinili consisteci of 220 black üne drawings from

Snodgrass and Vuderwart (1980). pnscnted on a white background. Each pichue was

170 X 170 pixels in size. Four 50-picture iists, ad one 20-pichue Est, were created such

tht M o f the p i a r e s npwented manmade objects and the otha Wnpresented

Nonsaw-word syiIabIe decisions, 220 pronomceable nonsense-words were

created by changhg 1-3 letters h m words used m the sybble or aaimacy distractin8 task

desc~l'bed in Expairnent 1. Letters were changed such that the n d y created nonseme-

word was pronomdide. A research sssistuit Pnd myselfverified the numba o f sytfabtes

in each word when read aloud. Four lists of 50 nonsense-words, and one iist of 20

noaseasbwords were constructeci. Each of these lists wnsîsted of one, two and three

syllabIe nonsense-words; Wof the n o ~ w o r d s fbr each üst had two syWIes, and

the rest had one or three syUaôles.

Ferbahdispne(iag*-arab--iak-ofic

cornpita smen at 1 rate of one item every 2 secoQds. For the pi-anllaacy ta&,

participants indicated if the presented pichire repmented a manarade object by makhg a

keypnss ushg their dominsnt writing hPad For the noacimsaword syUabIe ta&

participaats indicad ifthe presented mnsenraword hd two syiiables by making a

mm* ctask Asmbrpaimait 1, thestudyphPefwtbetargctrnemorytask

was aiways followed immediately by an arithmetic task to elimihate recency (as in Craik et

ai.. 19%).

lb2dm

session, Participants were tested individdy. As in Expairnent 1,

prrt*cipuits pafomed the target memory task uada ni1l attention, foIiowed by a practice

session on the pictute-uiimacy and aownseword syüabIe ta& The order of practice

block for the &trachg tasks was counterbalanced.

sessi~ng. The procedure was identicai to tiuu d e s a i for

Experiment 1, except t h t the Muiiacy and syüable tuks were reploced by the picturr-

animacy and nonsense-word syIlabte distractug tadis.

Asintbe~Eiq>erllnent,Iwisbed~,dctamiiaewhethathedistractug~

were equPlly dBcu.it and/or rrsource danandinp. Again 1 considered each participant's

paformance on the auditory CRT tpp4 p a f o d dom and in duai-task conditions wah

eaehoftbedistnctingtds. Themetbodwasthesrmeasmtrpallnem Iexceptthatthe

Results

Tamet memorv The nonsense-word syUab1e ta& intafend substantially with

h e & pexfionnmce- Howeva, fke d did not deciine as much wtsen the picture-

m h c y task was paformed comamdy at ntrievil The means for each condition an

presented in Tabie 3. The &ta w a e andyzed in t two between (order ofexperimeatal

condition and orda of singietask measure for the picture and nonsense-word task) and

one within (arpaimaatal condition) ANOVA. Thae were no signifiant main dkcts or

intemctions with the orda Won on h e nall paformance. Wts were sisnifiant at

p < .O01 uniess otherwise noted.

Thae was a main effkct ofexpezheatal condition E (2.46) = 18.89, M a = 1.94.

For each DA condition, the numba ofwords r d 4 was sigdicamiy lower than the fidi

attention budine condition, E (1.23) = 11.74, MSE = 4.35, P( -01 and E (1,Z) = 30.9,

MSE = 4.69 for tk picture and nonsense-word coaditions respectively. The mean

mimba ofwords d e d in the picture DA con&tion differed signifiamtly fiom the mean

nonr#nsawordDA condiaon, F (T, 23) = 9.2. MSE = 16f, pc -01. The diffaence in

mean peccentage decline in memory ( d o ofpaforrnance for each participant imda fidi

and DA conditions), betweai the pi- and no-word DA coaditicjns, was

signScmt, t (22) = -2.89, p <O 1-.

Expaimnt 2 - Numba of Woids Rcalled, Acaarcy Rates and M o n T i (ii L milfi,eAl.r,ads) on D h m h g Tasks, d Pementage Dedine From Single to hial-tasic

Condition For Each Megsufe --

Measure and condition M SD

Targct memory task Words recaüed

Full Attention 7.54 DA@-aaimacy 6.08 DA nonsense-word 5.08

Percentage Decline DA pichue-animacy 16.00 24.60 DA nonsense-word 30.68 26.03

Distractingmk Acairacy-

Baseünepicture-animacy 9 5 -14 DA pichre-nnrCy .85 -14 BaseIine nonsense-word 3 3 -17 DA nonsense-word .57 -18 Filier .93 -15

Reaction tirne Bsiaiacprairainim4cy 653-42 11235 DA pcRue-enimPcy 817,ûû 160.5 1 Baseüne nonsense-word 1073.54 195.64 DA nonsense-word 1062-29 198.68

P~eincmise iarrediontÛne DA 18.45 14.61 DA nonsense-word 4-02 17.02

Noto: DA = divided attention

F?dbw&JdDAcoaditioqthe~~*tl#ehaaeeted

words h m the target ta& d e r full attention. As in the f h t Eicpallnnt, f i

participants recalled any additional worda The meui mun&r of additional words recaüed

a f k the 'pictumaallnacy' and 'nomemmmd syllable' DA conditions was ody 0.42

(SB.= -65) and 0.67 @DL= 1.09) rrspectRnty.

Acauscy rates (dcuiaîed as #hW15 - # f i h alrnns/lS) on the nonsense-word

and picture-raimacy tsslg under DA conditions wàh âee recaii, were worse than singie-

ta& performance. The pacentage decline in acamcy rate, &om single to dual-task

conditions, was much iarger on the nonsensaword than picture-animrcy tasic. The data

wsn a d y d h a two h ~ ~ e e ~ (order of exphenta1 condition and ordr of single-task

measun for the pictut and no-word 6) and oae withlli (crrpaimental condition)

ANoVA Thac was a main e6êct of arpamiental condition, E (3.69) = 4 0 , MSE =

0.02. Mean acnuacy tates for each ta& in each condition, are presented in Table 3.

There were no significant main effects or interactions with the order fàctors on distracting

ttsirperformance.

Acauacy rites on the picture aad mnsense-word dUaoctiag tasks, under DA

conditions with âee rem& w a e worse than singiectask @ormence. (E (1,23) = 7.33,

MSE = 0.03, p < -05, and E (1,23) = 70.79, MSE = -02, respedvedy). Penentage decline

h s c a r n c y r a t c , ~ m ~ g l e t o d ~ t . s L p a f o ~ w s s s i g n i s c a a t i y ~ e r o a t b e

noasease-word thai p i c t m d m q task ('& (22) = -3.02, p < -01).

Rc8dmFiL

The mean RT to make a response, for each &trachg ciidg in the badine and

dual-task codtïon with retrievai 0 noted in Table 3. The baseiine RT was sigdcantly

longer for the n o ~ w o ~ d thoa pi- distmhg ta& ($ (22) = - 1 1.47). The

pacentaee maeaK in RT, Rom budme to duai-task conditions was sisnificantly greater

for the than noiisensbword task @ (22) = 4.34).

Tht coltelafion between RT on the dbtrxhg tps4 Md memory interfererice

unda each DA condition was non-signincant.

of?! CRT

stnctinn task. The accuncy rate was hi* on the pichire than nonsense-word

distrcictiag Eadg when each was pafonned concurrently with the auditory CRT task, t

(23) = 4.42. Thae was no &éct af& order on accu- rates. The mean accuracy

rates on the picture-MUnocy and noasansoword tasics, wae .67 GD.= -16) and .48

(S&= -18) respeaiveiy.

CRT tone taskt The Werence in the awnber of tones correctiy identifieci in the

picture and nonsense-word DA conditions was not sigdamt. The mean numbex of

correct responses for each condition is ptesented in Table 4. A withh participant

ANûVA reveekd a mPin &éct of condition F (2,46) = 107.95, MSE = 60.17. h

waa no main &&ts or urteractiow with the orûer fàctor. PPlaaned cornparisons showed

the numba of tones comectty identifid in th pidure and nomeme-word DA conditions

difked signifimtiy fiam the fiin attention badine condition, F (1.23) = 116.42, MSE =

152.61 and E (1.23) = 129.27, MSE = 162.94 mpctively.

Fjlpaimnt 2 - Num&er of Co& Responses, Reaction Times (ii Milliseconds) f ~ r Comct Responses on the Coatiauou Reaction Time Ta& for Each Condition

The mean RT to identify tones is shown for c o r n responses only (sa Tobie 4). An

outlier Mllysis e h b t e d RTs greater or lesser ihui two standard deviations from the

mean for each participant in each condition A within participant ANOVA revealed a

main &kt of condition F (2,46) = 18-54 M s = 13855.26. Thcn were no signifiant

main e f W s or intetactions with the order b o r .

The mean RT in both the pictue md nonsense-word DA condition diffend

sipmficantiy fiom the mean in the baseline condition, E (1.23) = 16.20, MSE = 21787.87,

p < .O 1, and F (1,23) = 2 1.00, MSE = 48523.42, rrspedively- The Merence in RT

between the two DA conditions was signifiant, E (1.23) = 13.46, MSE = 12820.27, p <

.or.

Discussion

A hrge interfèrence &kt on memory paformance, comparable in size to tbat

f d in 1, was obsaval unàer DA at retrievai ushg the wosmse-word

syllable tnrlc The pichm+ammICy ta& paformed conaaraiily with retriewai, produced a

deaakr-obv-ie-*- These--*

phoaoIogicai proccSgne of word-lüct materiel in a distrPaiag tasL is dcient to Unpsir

the reûieval processprocess In contrast, Kmuitic processhg of i tem in the di- task does

not mat neuty as great an m t e r f ' i m affect at retrieval, These results are consistent

with the suggestion h m Expaimait 1 that the prkmy lonw of interference unda DA at

retrievai ocws at a prasemantic lerd Tbat is, competition at the l d of word-form

n p h o t l s is d u e n t to h t e r f b with retnevd- These hdïngs are dso consistent

with the hypothesis that DA effects at retried ocnv primariEy due to cornpetition for a

representationai system (Fernandes & Moscovitch, 2ûûû), though they do not arclude

other sources of interference.

It shouid be noteû that single-task pediormance of the nonsaise-word task, as weli

as the pacmbiee dedine in acamcy rate h m angle to duai-task conditions is

compara&le to that observeci for the anbacy and syhble distrocting tssks Born

m e n t 1. Moreover, the relative iucrease in RT to i d e tones on the auditory

CRT tisk, fiom single to dual-task conditions, Y simüar in the nonsense-word task to

distraciing tuLs in Eqeriment 1. This argues a p h t the possi'bility that the nonsense-

word ta& creaîed interference at retritd mie to mutrsed tasfr-dEffidty, or resource

Even hugh the picture-aaimacy distioctmg Eppk did not produce as iarge of aa

intafaence &ect on wmory, it was noaetheiess signifiant. My interpretation is that this

resuIt argues a g a b t the corrclusion th.t memory hmfèmce ocnns goIeiy due to

tompetition rt the levei of word-form repmenîations. Semantic pmesskg in a

c l m m h g t i d r d n t h r r ~ e h ~ ~ d i g l l p t S ~ n t i r i e v r t . thougk

to a lessa degret. Another poss'ble i n t a p d o n is tbat the pictures may on occasion

advate the word-fom system Because this does aot oeair on aü nials, the interference

effèctisdimkhed.

It should ais0 be notd that the pichireanllil.cy tasic was shown, on multiple

mrairer, to be #ria to paform than the nonsense-word ta& Thae wa, a much d e r

decremeat in icnincy rate, fiom singîe to duat-tatk conditions, on the picture-inimPcy

task thn nonsense-word task. Also, when perfîormed ueda duai-task conditions with the

auditory CRT tasic, the aawacy rate on the pictweanimncy taslr was much higher, and

the RT much shorts than in the nonsense-word task It is possible thaî had the pichue-

task ban more dBiCUIt to perfonn, or more fesoulce-demanding, Iarger interfierence

& Î on memory wouM have been obser~ed Thus semantic processing in a distrecting

ta& might contniute more to intafaence edfiécts on mamory retrievai than m e n t 2

w d d suggest. 1 explore this poss1iüity in E q e r b n t 3.

Examination of distrecting ta& costs shows the expected decliae in a c a ~ a c y rate

fiom single to d d t a s k oonditions for both the pidure-.aimaCy d nonsense-word tasic.

Thy as previous work suggests ( C d et ai., 19%), the r e t r i d process is reSOUTce

danrnding as indexed by p r e r perfiormance when diptracton is con- with

rraievai. The sïze of the decrement in rcaarcy on the nonsense-word tssk Y gmilrr to

tbat observeci in previous studies of DA at retrïewd ad to that observeci with the

dutrpctllietasksfhmExpaimait1. Thedacnmaitonthepicture-sallil.cytsslris

d e r *

F ~ ~ ~ ( 2 0 0 8 > ~ * d i s t n a m g t a s k e o s l s * m t

d i f f e r d ~ o n t h e m a t d u s c d m t h e ~ ~ yet theyâid i n t h

apaimeat. However, just as the issue of trrL difiiaiity b k interpretaîion of the effècts

ofDA on rnemory, truk Miniity Wrdy phyed O mie in modulating the size ofdistractiag

tasic decrenents. The rrsults from the auditory CRT ta& showed that picture-uiimacy

wu xnuch aiaa to pedonn thin the aonraniaword Eidc This posgbîy accounts for the

d e r deaement in distndiag tapk pediormance uada DA codtions with r e t r i d .

Nonetheles, the sigdcant demement observeci in both disaa*ing tasks suggests tht the

major murce of interfierence, of the memory task on the distracters, W cornpetition for

general fesources rather than for material-specific systenis or processes. 1 disaus this

issue fiatha in the Gaiaal Discussion ofCbpter 2, p&r considering distracth8 task

c o s t r a a o u d t h r e e ~ i n t h i s s e c t i ~

Expcriment 3

Introduction

Even though the picnife-aiiimacy distractmg îask was much casier than the

nonsense-word the &ect it pmduced on memory was sisnificant. Theae d t s

pmeat me from coaciudïng that wmory Merence ocairs exdusiveiy fkom

cornpetition rt the l d ofp&oaologîcal or word-form npnseaEati011~. S d c

~ h t h e ~ c t u t e - t a i m a e y t i u L o b v i o u s t y k s o m c & a o n a r e m o r y r e t n ~

though 2 does not allow me to determine the adenî. The purpose of the

present aqKmnent wur to examine w h d m the magdde of the eaed of a pictumbasecl

chtrac@ task on meaiory wouîd hcmse, and a p p d that obsaved with the

~ ~ ~ W ~ ~ w P u ( r d . k ~ o f d i f e a J I y ~ ~

deniandS. h E x p a U n e n t 3 , I c r e s t e d a n e w @ ~ ~ ~ w h i c h n q u i r r d

participants to make size decisions about pi-es raîher than 8nimacy decisions. The task

hvoiveci identifying linedrawings that represented objexîs in the rd wodd that wae

bigger than an average cornputer momtor. A piiot stdy showed that the size judganait

taskwasmore~cu(ttopafomthanthepi~uiimicytask

In this m e n t 1 dso induda? the a o ~ d DA condition in an attempt

to repücate the resuits fkom th condition in Expairnent 2. The corollary prediction nom

F d e s d Moscovitch (2000) is that ifinterfefence &ccts at re t r id are material-

spcàfic, then mr, a d i fh i t pichrn-based diirtncring ta& shouid produce l e s

i n t e r f i on word retrieval than the nonseme-word syUable tasic.

In addition, aeatiag a more difficuit pi-based distracthg task aiiowed me to

re-examhtheclaimthatdecrementsindistracthgtpJr.cw~~cyrate, pediodunder

DA conditions with rrtneval, depend more on g d ~~so~rce-competition b e e n the

memory and distractllig task than on cornpetition f9r niaterial-qxciûc representational

Partjci0!xmu

Participants were 243 niBre undergmduate saidents at the University of Toronto

who recehd S10.00 for th& participation. An participants claimed to be native EngIish

speakers, and to have nomial or CO& to normai vision and W g .

Exparnient 3 was i d d d to Expairnent 2 eccept that the pictureanimscy

âbtndng task was repiaceâ by a ta& that required azt decisions to picnires.

Matallls

Tamet r d task. Stimuli for the target memory tasks were identic81 to those

used in Exparimats 1 and 2.

Picture-size decssions. . . Stimuii coiisisted of 220 Snoâgrass and Vanderwart (1980)

biack line dtawings, presented on a white background* Each pictute was 170 X 170

pixels. Four 50-pichne lists, and one 20-picû~e Est wae creued W o f the pictures in

each lin r e p m t e d objects tbat in the reai wodd wae bigger than a cornputer monitor in

ske, d t h e otna batfrepresenfeb objects d e r chsn a monitor in size.

Nonsaue-word svilabte dPcimom. The same mat* were used as in Expecllnent

R r w d i s c n e c i i i g t r s h i - - ~ v i a u l l y 0 ~ - & ~ 8 ~ ~

at a rate of one itan mry 2 secondsnds For the pietun-8ze participants Uidicated if

the prweatcd pictura npnsented an object m the ml worM that was bigger thrn a

amputer monitor in size, by making a keypress thm dominant wntmg b d . For

the nonsense-word syllable ta& participants idcated ifthe presented nopsawaword had

two @ables by &g a keypnss.

Anthmetic taskg As in Exprhmt 1 anci 2, the sftrdy phase fbr the target mernory

task wu aiways foloweû immediateiy by an a c tasic to diminate recency (as in

CraiL a d., 1996).

Procedwe

sessioa Participants were tested individdy. As in EXperiment 2,

participants pCnotmed the twget memory uak under fûii attention, foiiowed by a practice

session on the picture-ske and nonsense-word syilable task. The order of practice block

for the distncting tasks was counterbalanced.

sessions. The procedure was identicai to that describeci for

Expaimeiit 1 (and 21, arcept that the mirmcy and @able tasks wae replaced by the

picture-size md nonsense-word syilabre dlserctug uska

ComDarjn~ difficuitv of the diaractmn tasks,

As in Experiments 1 and 2,1 wished to detemine whether the distracting tesks

wae @y resource demandhg. These resuits were espedly important g b n that the

ai1nofExpimS3 wasto aesteapicbas-bPseddimEasktbatwrs~@Ieh

resource dcrmads to the nonsense-word ta&. I considered each participaut's paformence

are presented in Table 5. The data w a e anai- in a two between (order of acpaimental

condiion ud order of single-& masurr for the pichm and nonsense-word tyk) end

one witblli (expaimentil condition) ANOVA Rem,& were signifiant at p < .O01 unless

otherwise noted.

There were no signifiant main effects with the orda fistors on fne recaii

pafomance, though the Orda X Coadition interaction was signifiant F (1936) = 2.90,

MSE = 2.05, p<.Ol . For di orders, uie number ofwords r d e d in the fidi attention

condition e x d e c i that in the nonsense-word DA condition. For two of the Sa orders (ii *

which the nonsmsawordDA concRiCon was fien as the I* or 2? a < p e r l l n d

coidition), the IUrrnber of wotds reuciîed in the nonsense-word DA condition aceeded

tht in the pictiubsize DA condition by 1 to 1.75 words. For mther one of the orciers,

the number of words r d e d in the nonsense-word DA condition was 0.5 above that in

the pictumsize DA condition. For the nmeimag three orders, participants recaiied on

average 1.5 to 3.5 more words in the picaire-size tban nonsense-word DA condition.

TIlawasaopPEtiailrrocderof- EnriRitinnsthatd-&*Order

X Condition i n t d o a ; the htersdion appears to stem fiom the hi& variabüity in

mimber of words d e d in the nonsense-word DA condition.

There was a main Séct of acpaimeat.l condition E (2,46) = 13.66, MSE = 2.89-

For erdi DA conditions, the aumba of words recaüed wu sisnificantiy Iowa than the fùU

attention bad ine m a o n , E (423) = 13.58, M a = 5.41, F.01 and E (1,23) = 28.75,

MSE = 5.22 for the piotUre-size and ~ w o d conditions, respectively. I conducted

om-tsüeû t-tasts to compare the two DA conditions to cach otha since 1 had a directionai

hypothesis. The difference in number of words r d e d m the picturesize compareci to

nonsense-word DA condition qproached, but did not nach sta&icai signincmce, t (22)

= 1.42, p = 0.1 1. The merence in percentage decline in memory (fiom fidi attention to

DA conditions), for the picnire-size ad nonsense-word DA conditions, w u signifiant, $

(22) = -1.75, p < .OS.

Exparimat3 -NumbaofWordsRecalled,AaMcyRites.nd~onTimes~m mtlli4orrrnds) on D h t m d q Tasks, Md PercenSage Declnie From Single to Duai-task Condition For Each Measure

Measure and condition M SD

Targetmmiorytaslr Words d e d

Full Attention 9-08 DA pictumsize 7-33 DA nonsense-word 6.58

Percentage Decüne DA pictun-size 17.09 26.62 DA nonsense-word 28.47 26.0 1

Di- task A==w me

Baseh pidure-size DA pidure-sis B a s e h nonsense-word DA nonsenseword Fier

Reaction t h e Besdiee*& DA picturesize Basdine nonsense-word DA nonsem+word

FoltoamigdDk~tbepatiQpiiilsmre~iCht~toreeaIl

words fiom tba met ta& under Mi attention. As in tbe other Experknh fèw

participssts d e d uy additiod.words. The mtui mimba of additional words d e d

&a the pichiraske and nonsensbword syUable DA conditions was .38 m.= .58) and

0.89 (S.D== 1 .OS) respectively. Tbese were rot sismficady diffkent h m each other.

A4zwmmk

Aeairacy rates (caiailPted as W l S - # füse aiamisll5) on the noasaubword

and picturesize tosks, under DA conditions with fiee recaU, wera worse than single-task

@ormance. The percmtege decüne m .caincy rate, from sin& to duai-task conditions,

was larger on the nonsense-wod t h picbire-size task. The data were d y z e d in a two

berween (ordn ufarperimental condition and order o f singfe-task measure for the pichire

and nomense-word task) and one withni (acpaimeatal condition) ANOVA. There was a

rmUi Héct of- condition, E (3.69) = 30.03, MSE =0.02. Mean acauacy

rates for eoch ta& in each condition, are presented in Table 5. There were no signifiant

main &ects or interactions with the order fàctors on distracthg task pafonnance.

Auniracy rates on the picûue and noasense-word disaactuig tasks, unda DA

conditions with âee recaii, were worse tben singk-task @ormance. (E (1.23) = 36.38,

MSE = 0.02, anci E (1,23) = 5 1-54, MSE = .M, rapddy ) . Pefcentage dedine in

~wrarcy,hm~todd~performance,wusignincenrlylorgerontbeao~

word tban pidurr-size task @ (22) = -2.34, p < -05).

R e d o n n~~leL

T h e ~ t i m e c o ~ ~ i t e a i e s ~ b o s k , k ~ ~ d d w C C s r k

coadition with rietrievai is noted m Tabk 5. The beseline RT was sienificady longer for

the nomemeword than picture-size dhtmthg tDslr & (22) = -7.6 1). The percentage

inaeuit in RT, h m b;isdint to dual-task conditions did mt difila significantiy for the

picture-size and nonsense-word Easlc Q (22) = 0.93, p > .36).

The corniotion between RT on the distioctiog ta& and memory interfice

uida eaeh DA condition wu non-si@cant.

OW CRT

g taskL The acauacy rote for both of the distracthg tasks dered to a

simüar de- when the auditory CRT tasic wu perforrned conmentiy. The mean

acavacy rates on the picture-size ad noasaise-word tasks, were -54 m.= -15) and .55

(S.DL= .29) rrspedivdy. There wu a mrrginaYr sigaisant &kt oftask order on

acairscy rate for the picture-size task E (1,22) = 6.90, MSE = -13, p<.O5, with those

paforming the task as the fmt dd-task condition with the CRT task, v a s u second,

obtainllig a mean sccuracy of .47 (SD. = .l3) and -63 (m. = .29) respectively. There

was no &kt of ta& orda on ecaincy rate for the noasaisaword task.

mtone The d i f f e r e ~ ~ i e Ui tk mmda oftones c o d y idenfidied ùi the

pi-size d nonsease-word DA conditions was not si@cmt. The mean number of

correst respoaises for each condition is preseated in Table 6. A within participant

ANOVA revealed a main &ect of condition E (2,46) = 72.76, MSE = 95.86. Thae were

no msin effear ofthe order h r . Pianned oompPrisons showed the numba of t o m

c o d y idedieci in the pibure-size rad n o ~ w o r d DA conditions diffefed

? i + a a d y b k ê a L o t t ~ b î s e l i n s ~ ~ ( 1 , a)= 77.- A-S-=25Â%

and t (1.23) = 117.40, MSE = 189.65 respectmly. The <iumba oftones identified in the

pictuiasizc versus nonsense-word DA conditioa wu not signifidy different.

E>rpaimait 3 - Numbar of Correct Reaponses, M o n Taies (in Milliseconde) for Correct Responses on the Continuow Reaction Tmie Tasic for Each Condition

C o m a rrspoiise Reaction time

The mean RT to identify tones is shown for correct nsponses oniy (we TaMe 6). An

outlier analysis diminaieci RTs greater or lesser than two staadard deviations fiom the

mean for each participant in each condition. A withïn participant ANOVA revealed a

main &ect of condition E (2,46) = 43.15, MSE = 8505.95. T'hem were no signifïcant

main & i or interactions with the order Wor.

The mean RT in both the pichme-size and nonraise-word DA condition d i n d

si@cady nom the mean in the basdiae condition, E (1.23) = 54.84, MSE = 20652.66,

and E (1.23) = 61.99, MSE = 17239.96 respedvdy. The Tbe in RT between the

two DA conditions wu not signincady diffZreat

k m of v e n t C O ~ D M 2md3

c&ct of the picture-bkced diatnctiag ta& on memory, and vice versa, 1 re-analyzeà the

datafrwa~Zaad3,usinganANOVAwithE><pallaaitasabetween

participant fktor. 1 aiso wanted to detamine wheiha the interference akct on memory

ud the distncting ta& was similu in the aoaseasaword DA condition, across

ex pi ma^ 2 and 3. A separate aniiysis wu, d e d out for the pacentane d e c h in

d u d d i a t s d r a c a i t a c y .

The main &kt of Expairnent on percent decliae in manory task @ormance was

not sisaificant, E (1.46) = 0.01. The main e&ct of Condition was significant, E (1.46) =

9.94, MSE = 0.04, but the Condition X Expairnent interaction was not, E (1,46) = 0.16.

Planned t-tests showed thot the B e d e c b m rnemory perfiormance did not differ

in the pi- and picturesk D A conditions, g (46) = -0.15. SWW, the

pacaityle dedine in memory did not di&l in the nonsense-word DA condition, across

Expehem 2 and 3, t(46) = 0.29.

The min &ect ofExperiment on permtage d&e in distraaing task

performance, was wt signifiant, E (1,46) = 3.20. The main efféct of Condition was

sipificant., E (1,46) = 14.52, MSE = 0.06, but the Condition X Eqehen t intaaction

was not, F (1.46) = 1 S8. Plamied t-tests showed that the percentage dedine in

distnctgie tuk pediormuice did mt Mer m the picture-rnmiacy and pictwe-size DA

conditions, though the difference appt08Ched sipificonce. & (34) = -1.87, p = .07. The

perœn&ge dedme in distractirg task performance did mt différ m the waseose-word DA

c o ~ o n , n s s E . 2 rnd 3, t (46) = -0.19.

A ~ U i t c r f e r e n c e & ~ o n m e m ~ r y ~ ~ ~ o b s a v e d u n d a D A i t

d e v a i usiqethe no11setls6.wotd tasic. The &bct was comparable in oize to

tbat f d in Expriment 1, and feplicated the tesuhr h m the nonsense-word DA

conditon of m e n t 2. The picture-size ta& pduced a d e r e&d on memory

tbough it was dl signifiant.

Redts h m tiae auditory CRT atk show thrt the picturosia lrnd nonsense-word

distncting tasks were e q d y resodemanduig The n u m k of tones identifieci in

each dd-task condition, and the RT on the auditory CRT task did not Mer signincady.

Moreover, the acauacy rates in duai-task conditions with the auditory CRT task were

smiüar for the picture-size and no-word distractuig tasks.

Thuc, despite mrlaiig the picbin-bascd distncting task more resource dem~ding

it did wt produœ a iarger e&ct on memory than the picturMaimacy task nom

Expaimait 2. As in Expallneat 2, the ~ m ~ w o r d DA condition produced the iarger

interfèrence &kt, suggesîing that orthographie or phom1ogicai procesptig of word-like

material in a disaactùig task is dcient to impair the retrieygj process.

The tmnmw+word DA condition prodiiceda 6 @ d y mer percentage

dcciïne in mmory paforrnaaœ ( h m fûJi attention) than the pichire-size DA condition,

though in absolute terms (mean munber of words tecalleci m each DA condition), the

~erencebetweenthandidnot~signaicamr. ComparisonbnweenExpaimeat2

and3 s b o ~ t h t ~ p i ~ s i z e t L p L d i d a o t p ~ a ~ e r m O a f a m c e & ~ t h s n t h e

p i ~ ~ t r d g u d t h a t t b e ~ w o r d t r J F i a b o t h a p c c i m a i t r p r o d ~ ~

~0fvsrysi imlrrBze.

nie larger intafaeace effkct producecl by the no~1~etl~e-word compareci to

picture-size dkmdng task niggests that the prireaor locus of interfierence unda DA at

retrieval ocam at a pmdc M. This does mt, however, ruie out the possii i

tbat tbae may k 0th- ways of intedixhg with memory retnevai. 1 discuss this issue in

more detoü in the Gencnl Disaidon of C-er 2.

I n d o f m y e x p a l l n a i t s I ~ r e ~ p ~ ~ l ~ e R T o n t h e ~ g t r u l u g n C e I

ini- balieved it could provide an indirect measure of resource-reqwtenients for each

distract@ task. However, in Eqxhent 1. the bapdine RT on the animecy and @able

didnctiag tasks d i f f i as did the percentage inatiae in RT d e r DA conditions, yet

uiese trdrs produced Smüar amouna ofhterfeteace on manory. This suggesu that

différences in fesponse RTs, on the -or tada, are not relateâ to the mwt of

interference ptoduced by these tests on memory d e r DA conditions. Moreover. the

d t s of the Iinahpis of respoase RT were o f h in the opposite dinaion to those found

on the auditory CRT ta& posgily beause 1 emphsind rcnincy in responses on the

distracthg tasks, a d did not exptiütly infonn participants thDt RT was also king

measured. Thus, RT ca~not be considered as smdbe a measme of task-diffiwity and/or

resocna d d as the auditory CRT ta!&, which was conducteci expliatiy tu evaiuate

this issue. Also, the auditory CRT ta& meaaaes pecformance of each distracting tasic

independent of its effkcts imda DA with memory tetriWevaL As such, 1 believe it provides

arnorea#mrtemua~eoftask~adtyrndlorn#wcenquinmeaSsforerch

d i s a o c t m g e

Even though the pictme-size and nonsense-word tas&s wae showri (on the

auditory CRT tasic) to be eqq@ dificult or resource danondiag the magnitude of

interfèma oôsared on epch task's s m a c y r a ~ unda duai-task conditions with

ietnevai were diffaait. DA led to a ZW!% and 33% deciine in distractuig task acairacy

me for each ta& mspthely. These Thseth repnwmt f@iftcant d e c h in d~stractbg

ta& paform%nce, and thus support the idea thrt reüievaf U rrrance d d g ( C d et

al., 1996; Jobnston et ai., 1970). However, the diierence in magnitude of the decrement

to the dUtracting tasks suggests that in addition to germai resoufce cornpaition, thae

may be a material-Speanc component to the &'kt. This issue is addressed in the G e n d

Discussion of Chapta 2.

Ownll, the nonsense-word distracting tuk produced large interference dfiects on

memory, simüPr in magnitude to those observeci in Expriment 1 and to those obsaved

âom vabal-based distracting tasks in Fernandes and Moscovitch (2000). Thus,

phonologid proccssiag of word-forrns, without semantic content, can si@cantiy

disnipt the retrieval process. An equalky âifEcult p i c t m h a d distractkg tasic produces

signifiant, though d e r effects on memory, hdhtbg tnet saniatic pmcesshg play a

sxnaiier mie than word-form processiag m produ- iiitafbence at retrieval. This

suggests that even though d c pniasfoie in a distnctiag trsk is capable of disnipting

retiievai, t k a o t t b e k e y ~ o r ~ t h e 9 t e o f m m o r y i a t a f a a i c e f r o m D A c i t

retrievaj*

scnriC-&w%-

The present ex@mats d m hge Mterference effects on recali of a üst of

unrelafed words, w&n a distrachg tasic invoIving word, or word-like, material is

pedormed corrcurrentiy. The msgaibide ofthese &CCtS are 9mürr to those nported by

Faii.ades and Moscovitch (ZOO), w h a vabrl mmhg recognition ta& and a word-

momtoriag distndng tasic Ied to an appmxhate 3Ph deciine in memory pedionnence.

h m fÙü to DA conditions at retrid. These resrrits stand in contrast to other studies of

verbal wmary in which retrieval wss p a f o d conamently with non-verbai distracting

tP1J(S. such as card-sorting (Baddeley et ai., 1984), digit-monitoring (Fernandes and

Moscovitch, 2000), or a visuo-spatial tasic ( C d et al., 19%; Naveh-Benjamin et al.,

1998), which kd to smaii interference effects on mcmory.

eCan Ta& Performance

niis series of acpaiments was designd to ùnrestjgate which as- of verbal-

b d d i d g tasks lerds to large htderence effects on memory, and in doing so

provide insight into the naouras. and undedying neurai systems for r e t r i d . 1

imtlllty hvestigated whether it was necesocuy to have a memory cornpanent in the

dbtradng task to produce mtafuence. An of the v a dùtractmg tadis âom Fernandes

and Moscovitch (2000) had a memory lad, which la open the possii'bÜity that

in te rkme eS&s coufd poteatially k due to wmpetition for memory structures rather

than for a common, wotd-basexi representationat system as they suggested.

Intbenrsta<paimenfdespite~tbemamo~compoaaitofthe

d h a d q ta& 1 f d that oa-line word-based cisLs such as anhacy or sybble

jadeeaiaiits---*&m~-lsFg6intatfiB€eRœ&i

Firrthern#,te,thweeffectswaenotmoduhtcdbytbalevdofpdgrequindh~

distncting ta& wbetba semantic or phodogid This suggested tbat the word-f0rm or

phonologial aspect of the words in the di- ta& ntha thon their semantic content,

accounted for the hrge intdereace &kt on memory.

Consisteni with this suggestion, Exprhents 2 and 3 showed that even a

dismctiag ta& involring nonseme-wonfs, that bss m semmtic conteat, produced

comparabiy iarge effects on memory. On the other band, a s d c d i m g task

imroiving pictures with no word-form component, pduced a signifiant, but smaller,

&ixt on rnanory (Sec Figure 1). Inmas@ the I d of difnaiity of the picture-baseci

ta& h m Experiment 2 to Experiment 3, did not in- the size ofthe interference

&kt signaiîsntly. Thus, to have the largest effèct on mtrievai, the task shouid

engage word-fonn or phonologid processes.

DA non- animacy syllabie pidure- word pidure- word

animecy Expt 2 s ~ e Expt 3

Condition

1. Mean percentage dedine in fiec ncall paformance from full to divided attention @A) for each condition in Eqehxmt (Expt.) I,2, and 3.

Sime the nonsense-word ta& which did wt bave a semantic processing

component, lad to large interference &ects on memory, semantic p r o a s h g in a

distracting tark is not essential in disruptkg memory retnd. Consistent with this

suggestioa, the picturc-UllmpSr and pichmsize tasks, pafonned concurrently with

rrtrieval, produad significsntly d e r declines in memory (1647%), than the nonsense-

word tasks (28-3 1%) in E x p b m t s 2 and 3. Thou& d e r , these d e c b in memory

fiom the pi- tasks are nonetheless signifiant, suggesting that semantic processing or

cornpetition for generai resowces can also disnpt the retrievai process. Alternatively,

becaue the pictures wexe line drawiags of cammon abjects, they may have accesd the

name on some trials and i d to interfikencc &CCtS at a phom10gical kveL

It is possible that any distractllig ta& that is resoufce-dernmding disnipts the

retneval procea~ somewhat, due to the added complexity of coordinatiag two tasks. The

sizc of the interference Hêct produced by the pi-baseci tasks is not that diffaait fkom

whet has bem found in other shidies that used non-verbal discracriag tasks (Expiment 4

Faaudes & Moscovitch, 2000, Cr& et ai., 1996; Nav&Bajomin, 1998, Andexson, et

al., 199%). Thus non-verbai dktmchg tasks may produce d but reliable interference

effects on manory retrievai for an dtogehr diffaat reason t h the iarge e&as aeated

h m word-based distracting tasks. Pabrps some unouiit of atteation is required for

memory retrmPl and is comprom*sed by h h g to coordinite motha cedr dong with

mew,ry d e v a l , leading to d yet si@cant &eds of DA CoaSstact with the

~opsychotogicsl modd desaibed h Chapter 1, ecpéory ( W o n of the memory

tnce) stin be automstic. but post-ecphoric scPreh end post-ecphorï~ monitoring

F = - m i r r ~ ~ @ = b e f o \ i P k

Performaq -

A n o t b s s i m p o r t a a t ~ â o m ~ s h s d y i s t h r t ~ r a t e o n r l l o f t h

dktmdq urLs d e r e d under DA conditions with &d. Thus, in agreement with

Craür et ai. (1996). Andetson et aL (1998) and Namh-Benjamh et ai. (1998), retrieval U

not an ~ornatic procesa, as it draws remrces away h m the dUtractiag taslr unda D&

Ieduig to d ï s t m h g tiisk costs. In tbis series of the magnitude of the

decranait in acauacy was mss #paimaits. Figure 2 shows the percentage

decline in wiracy for each task in each expriment.

animacy syilable p iduP word pidure- word animacy Expt 2 size Wt3

Condition

1 attnie the very d deciine in the picturaimmacy ta& to the fia that it was

shown on dt ipie messuns to be 1- dificuit and less resowce-dernanding than the other

task, in this study (sa 2 Results and DWnwsion). Looking at the

performance costs for the raneinllig distrading tasks, notice that the word-8nimacy task

fiom Experhmt 1 and the pich~osize tssL h m E;rtpaimat 3 show a datively d e r

deciine m paformance than the otha tasks. The wwbriiimocy tPrlc pioduced large

effects on mernory, similar in size to the syUable and nonsaise-word distracting tasks,

despite the fact thM its accuracy rate was not as affectecl. The picture-sUe tssk produced

d effécts on mernory? but its a c u ü ü rate declined by the same amount as the word-

aamiacy tsslg which produced Iarger etfécts on rnaaory. 'Riese r d t s suggest that

chtmthg ta& costs can differ depending on the tasic useâ to divide attention during

retrievai. However? material-speciticity of the distrrctiag task, pedionned comurently

with fkee rrcall, is not the faaor produchg this variaôtlity. Ifit were, the word-aaimacy

ud p i c h ~ m i z e distractirg tasks wodd have Ied to diffkrent amounts of decline, and wt

the same amount, as 1 have found.

Tfius, ifTconsider paforrnance CO- 8SSOCiBtedarithDAmssall three

acperiwn~, these fesults support the dami thrt cosrp ocnir independently of material

useû in the d h a c h g task (Fernandes & Moscovitch, 2000). Whether these costs ore

a c t d y Urcocired with the maintenance of retrieval mode W o r reflect gaiaal fesource

rrqimemntr rcmnllis to k detennined. 1t L worthwbile to note tbrt the kgest e&ca on

distnctiDg tssL perfiomxmce were obsared wben the dkacthg ta& rtquirrd

p k o e d o e i a C m . ~ ~ a r t ~ ~ o n t b e m e m o r y a n d

distractin8 ta& may be due to cornpetition for phoncdogical pracesping resources.

A adous finding across aii of the e x p r h m ~ is that participants could not recali

many additional words fiom the shdy list &er the DA conditions were ova. A simüar

fincihg was reporteci by Feniandes and Moscovitch (2000). Tbey mggesteci that short-

tam consoiidatioa or cohesioa (Mo9coitch, 1995; N d & Moscovitch, 1997) may be

disnrpted by the cornent actiV8tion of mocortid representations needed for the v d a l

memory Md distrMing ta&. Given that the nonsense-word task was also able to disrupt

mamones, mn after the DA condition was complete, this aiggests tbat the locus of the

prolongeci interference &ect is basecl on accesrr to photbologicai or orthographic codes,

though the present acpeiiments do not d o w me to decide between these alternatives.

ConcIusions

These arperiments showed that d e r DA conditions with retrieval, distractuig

tasks n a d not have a manory or semantic procashg component in orda to m i t e large

interfefence effbcts with memory. This led me to augeest that large intafefe~~ce effects

ocan prhwiïy due to cornpetition af a pre-semantic tevd Consistent with thk

suggestion, 1 fouml that a nonsense-word distracthg ta& that had no semantic content,

prochrced I q e effkcts on rnemory. Thus, tompetition ot the levd of a word-fom system

is d c i e n t to produce large i n t e d m &eds on &ee ncirll; cornpetition is for

orthographic or phonologid npmmtations caanot be detaniined h m these

arpaimeats. An epuii4. âdficult pi- diftroctiqB ta& produced sipificatit

0 l l m f & w m a r i t 8 ~ ~ k f f h - l ~ ~ ~ & k ~ t h R t h P t

pmduced by using wordc or nonseme woids m the chtmdng trpL

1 lLpo fOUld that accmcy rate in the dhacthg tasic is aiways sipificantly

affccted whai performeâ connunntly with retxievai. The &tude of tbis interference

d e s for diffèrent distracthg tasks, and is unrelated to success at retrieval. Furthennoce,

tliis wiation cannot be uxounted for by type of Mtalll used in the distnctiag urlt. It

nmriDsto bedetamMdwiiethertbecuststothcdi~task ue8950Cillfeû withthe

maintenance of retrievai d e ( F e d e s & Moscovitch, 2000) W o r reflect general

naource requirements for reüîeval (Cr* et al., 1996).

By showing that tetriemi can be digupted by PafoRning a concurrent tnsk, these

resuks indime that r e t r i d does not always ocnn obügatorily. Fudmnore, these data

suggest thpt this disniption ocairs primariiy when there is cornpetition between the

distracthg and memocy task for wcess to a word-fom system which is naded to

m a c h t e the memory tnce, ad to pafonn the dhtm@ ta&

Cluptcca

In this chspter, 1 examine wbetha a n s o w d m d explpnatioa can ;iiccwnt better for the

pattern of M e r e n c e & i h m dual-taab at nhievai, thn the nâiropsychologicai d e l descrried

in the previous 2 chpters. The work by Fernandes ad Mo9~0vitch (2000) suggested that performing a

word-brsed ta& co&y with ncrll ofwordq intafixes with reactivation of ncocorticaî

npreseatations that are part of the memory trace. CompaitiOtl h aeated as the m m r y and distracting

taskscompcdefbrthe~npnsaitatiodsystem, fbedrtr~tedinChapter2arccoaeistmt

with this account of DA effects.

An aitemaîive account of these r d t s is tbaî the large mmory costs, fiom simüu materiai in a

di- is due to cornpetition for g d aiteutid resources, such as those mediatecl by the

âontai lobes. As mmtioned ù, Chapter 2, these fesources may be needed to coordinate the onüne

p r d n g of duai-tas&s. In the case of the word-based dktrdng tasks. retriewd moy be dimpted to a

dativeiy pater extent becswe the siniiknty in materiais makes it more dif6cuit to coordinate the two

t a d q anci overextends a limited pool o f g e n d proceshg resources.

For example, CRiL (1983; in press) has suggested that the ability to retrieve an item firom

memory am be likmed to the ease of pubg out a target h m tmoag diffhnt backgrounds or -

distradets, m p e n e p d acperllneata Thsf is, one's abinty to pick ciut the t q e t is eaSrerwbarttre

background offZrs a hi@ contrast as opposed to a low contnst. The aôility to fesoive m e t items is

nducedwbenitisrmongothermetaiJstJmtmebightysâmiiar~ Inordertopilloutatargetitem

~ ~ m f h î p ~ ~ y o l l t m ~ ~ t ~ a g r e a t e r c ~ o u a t o f n s o l ~ p w e r .

In tarmp of memory pafDrmance, resohg power can k mterpnt#l to wpn general

processing rcswrces. In th ase whae attention is dMded betmen a memory and distnctiag ta&, it

--- e u y k t h t ~ s t f i i f h u s c 9 m i l a t m a r a ; r i . ~ l p e r t c t ~ d m i p n d r t b a a ~ ~ u s e

dissimilar materialS. As attentionai ritsoufces are b h d , tbis resuhs in a break-down of the r e t r i d

proceas, accounting for the large i a t w effed o b d âom word-baseci distractllig tasks on

retrieval.

Anotba resom&wd account ofthe large material-qmific e&ct of DA at retrievai, is that

cornpetition b created for an inputdufpuf cfuiaad RoceaQne of incoming words for the disacaing

tasic mry t e e r vabal wot.king marory system, whüa at the nme time. words for the tedl task rnay

need the same mouces, W o n output. It is m üne with Badâeiey's (1986;1992) hypothesis that the

a b i i to coordinote connrmnt tasks relies on the caitral e x d v e (CE), whose operation reqWres

resource mediated by PFC, in his working mmory modd Coordination of conatmnt t~ sks h t

require the same sîave sub-system, in his modei, W more diflFicult than when tasks require the verbal and

visuo-spatiai sub-system respectively, becousc in the h m case they are relying on a cornmon pool of

resou~ces. Such an account suggests thst words to be d e d are SU-y re(LCfiVElfed, but then

disnipted during output. The wmponent process rnodd. on the otba ban4 suggests interference

oonus during repctivation of the memory traceq prior to output, and that int- is not b& on

cornpetition for gaKnl resources needed for PFC (and CE) hction.

O a e w a y o f f e S t i a g t h e r e s o u r c e ~ t o ~ - p r o c e s s ~ o f m t a f é r r a f t ~ isto

consider the paformance ofolder individuais unda DA conditions. It is wdl known thpt aghg is

rccompanied by a de& in long-term e p W c memry. A theoretical accoUItf of age-relateci decline in

memory i s t b t g d pPocessmg rrawrc*r tht are critical for l l t l l l l e f~ l~ cognitive opedom deciine

with age ( C e 1983; Cm& & Byrd, 1982; Raôhwiq Cr& & Ackefmat~ 1982). Theae resources

have aiso been coiicepuiüzed in tams of redwed working wmory cspoQty (Sabmse, 1996), neeûed

- cobelbd--~~de-(CRüc&- L%EWwF&Zscks,

1988).

Regardiess ofhow these t sources are coaccphiaüzad, wgritk aging theorists suggest that

changes m bmin M o n , partidady in the h n t a î lobes. d e r i i e the reduction in resouras (e.g.

Baddeky & Whn , 1988; Fuster, 1997; Knight & Graboweclry, & Sabini 1995; LuM, 1966; Shallice

& Burgess, 1991). Evidence of Merences m frontal lobe f h t i o n between young Md oM have been

s h m behaviomgY ~ o s c o v i f c h & W-, EH), and in s c v d t l cu ro -wg shidies. When

pdorrning a retrievai ta&, olda ad& show a reduction ni right PFC (and comsponding increase in

IeA PFC) compued to youag aduits, who othubit fiontai lobe activation primarüy on the right side

( C h a al., 1997; Maciden a al., 1999). n e r e is rlso some evidence of ligarelatad reductions in

cerebd volume (Coff9 et ai., l992), in regional d r a l Mood flow (Gur et 1,1987), and metaboiic

rates for oxygem Md glucose ('Leenders et PL, 1990; Pantano et ai., 19&1), partiailarty in the fiontal

lobe, though thon is much individuai varhtion.

tmew o f Chanter 3

A consideration of DA effects in olda aduhs dows one to detemllne whaher cornpetition for

generril pmceSgag resources, and the h n t d lobes, pIay a d e in mediating large interference & i at

retrieML As d e S c n i i n the component-pfocess modes the h n f a f robes are centraî syste5115 that

requin generai reswrces for th& opaations- When the bntal lobes deteriorate, they d m on an even

pater amwat of tesou~ces in order to carry out opaations &ecfivelyy If interfixeme âom DA occun

chat to a reduetion in aveürMe pmoesPing resoutces, thea performance ofthose with paor fiontai

nmction rbaild show amplified intafaaicc gnCe thy rsquin more nsourcer to mPinhin paf9mce

even underm attention.

I b v b u s a r o d r ~ t h ~ a f y ~ Pnd alda aduhs d e r DA conditions at

retneval f d tht Urùig 1 ~ ) ~ ~ dktra&g tasks hd liak e&a on memory (Anderson et al..

1998; MPcbt & Buschke, 1983; Nybag. Nüsson, Olofbon, & B(danan, 1997; Parlg Smith, Dudley, &

Lafkonza, 1989; W B ï g & Smith, 1997). Pafomisnce on the disercting ta& howeveq was dimpted

more in oida than yow~ger aduits (Anderson et d. 1998; Cr& & McDowd, 1987; Wbitllig & Smith,

1997). TbU Id Andasan et ai. to conc1ude tbat olda duks bave a reduction in resources adabte to

engage in damrndiiig opaotions, as uidexed by hi* dirtnctllig tasic costs, but relatively preserved

retrieval opsntions. That is, in agreement with Sdthowe, Rogan, & Prül(1984), olda aduits can

effectively &tain retrievaj d e r DA conditions, however, they do so at a relatively pater cost to

attentionai resources.

In tems of the component process modei, memory wu not disnrpted in these studies as

retneval on tests mediatecf primarüy by the MI'LlH pmceeds obügatorily. Perfomance on the

dishacting task was affecteci, under DA conditions, as this masure rdects the resourcedemanding

aspect of retrievai: estabüshing and mUntnining retrievaI mode, and monitoring output, aii of wbich are!

mediated by the PFC.

In this study, 1 tested the prediction b d on the component-pracesp mode1 that the large

ùnerfefetlce & i Born worcbb& ~ t a s k s i a d o e t o c o m p e t i ~ f o ~ eommoar-

in the posterior neocartex, rather than br g d murces, required by the âoatal lobes.

In-4 ofthisthegs Iuisessedniemoryinyoung andoldaddtsuadafiüland dMded

mention conditions. Because olda duits have de6cits m episdc memory, 1 sought to equate the

badinekvd ~ f W b e t w a n y ~ a a d o l d b y a ~ ~ ~ t h t o I d ~ o m i l t s t o üstentothestudyüst

twice instead ofjust once as in the young a&its. Ln 5 1 considered the p a f o m c e of

-- ~ ~ o f d d a ~ i n a r k i c k t l r ~ l i s t ~ ~ d y a r i c c InExprhw~Lfircthtr

evahiate the component-process d e i , in terms of possible brah regions mediahg the &bcts. Hae 1

examine @ormence unda DA conditions, in a population of d d a adub, classifieci by levei of

dydhction in tanpod mi hntai lobe M o n .

Expuiiitnt 4

Introductiotl

The question of interest for t& present arpaiment is wbnber the lugt mstetial-qedk

interfience tffécts documenteci in young adula (Fernandes & Moscovitch, 2000, submiaed) are due to

a reduction in geaenl processing resowces &om tbe DA condition, or to cornpetition for

reprrsaitatonai systems* To address this question I considered whaha older addts, with presumed

reduced ge3aaal procesPing resources would show interfaence effkcts, nom DA at retrievd, that were

Iaqer in magnitude than those obsaved in yomg aâults.

In the presait expehent I sought to d e t h the cffkcts of aghg on two différent types of

dimachg tas&. 1 cornpend ditectiy a digit tuk, which has ban sbown to b e r f i e M e with memory

for a list ofwords (Anderson ct al., 1998; CRÜC & McDowcf, 1987; Park et al., 1989; Whiting & Smith,

1997) in y- and old addts, and a word-based distractllig ta& whose interfiring &kts are

sub- h young ad* (see Expaimeat 1 of this thesis)-

As predicted by the component-process rnodel, I do not expect the siEe of the interference Hect

to di&r between y- amlofcl, whai eitha the word-rnimPcy or odddjgit d k a d q tssk is

CO- @id. Insofhr as iaterference acoirs on the memory it is due to competition for

arepnsariitiodoystem,wbichUrelativeiywdlprrsavedmthtoldasdul~ andwt dueto

w~onforgenarlpmassiiigrrsaaceriatbePFC,whichuecompmmised~thisgmup. A

r- p r m & ~ ~ & b ~ p c ~ t t u + - w i l L b e ~ i n o l d e r ~ k c r i r u

thar geaaal proceshg resources an reduced.

1t is posaile, bowever, thrit 01der adula mry nmemba ber items overaii, since episodic

memory decüaes with age (Cnür & J e 1992; Saltbouse, 1991). To equate b d e manory

pdoonnrace with young I presented the study word iist twice to older ad& iasred ofjust once as 1

did to the y- adults.

inadditi01ttot&~ofDAon111«~)rypaformarnce;~WBnfedtoaCaannehowcoststothe

distracting tasic wae affecteci by DA conditions with retrievai. To do so, I compareci the &ect of the

target memory task on i icau~cy and d o n time on the âistmdng tdcs. As predicted by both the

mmponent-process and reduced-resowce model, older duits are expected to show larger CO& on the

distractin6 tssk thn younga aduits; such COM are thought to be iacurred in maintainhg ntnevaî mode

(Anderson ct al., 1998; Craik et al., 1996; Fernandes & Moscovitch, 2000), a resourcedemanding

bction asc r i i to PFC, and which is compromiscd in olda ad&.

PamaDPnts

Participaa~ were 24' undergraduate students at the UMvasty of Toronto, who received course

cndit or monetaq cornpendon, m p e d d y , for participahg in the W. 24' seniors wae chosen

âom a pool of volunteas used for studies at Erindile Coiiege, rccnilfed by ads in the l d newspaper,

and rccQnd t o h r n o w compensation for puticipathg. The mean age was 19.54 = 1.14) for

the young md 71.69 Csp = 2.82) fw the otduadahsadahs The mcan numba of yecin of ducation f9r the

undergraduates was 14.2 (SD = 1.1) and for the older ad& was 14.5 = 1.9). An participants were

native Eaglish speakers, and had relatively n o c d or corrected to nonaal vision aud hearllig. The Mini-

Mentai State EXM (MMSE; Folstein, FoIstein, & McFIugh, 1975) was rdministaed to each olda

adult. The mean score on the MMSE was 27.7 (SP = 0.88). Participants with a score of kss than or

qua1 to 26/30 were repbced.

The produre for young and older aduhs wrs simiiar to that used in the previous experiments,

arcept that word-mîmaq and oddd0git identificati-on were the dUtraaing ta&.

For t&e olda ad&, the study words were played at a buder volume. The vohune was adjusteci

for each participant dürhg the practice phue, to a Iévef rt wKdi the order sddt coula riesr the words

cl@ without stninnig tbeir h&g Additidy, in orda to equate tmehe r e d between the age

groups, the shuty word üst was piayed twice during the q e r h e n t a i sassion for the older aduits.

O O deasions. The mterials used for î k distrachg tapk were the same as in the word-

aaunaçYtaskhmExpvimeat1.

s~ons. Four Sû-digit IjStS, and one 2O-digit E s t were created by choosing a string

of ruunbers, pseudo-domly, h m a table of mdom ambers. Each lia was crated such that halfof

the! numbas were odd and haifwere mn. Stimuli for the digit-monitoring task were two-digit numbers

chosai fiom a table of random nwnbers 1995). In the 'odd-digit' task, two-digit tlltmbem were

displayai on the oomputer anen nt a rate of one aumba every 2 seconds. Participants were told to

press a fcey, using the index finga of the dominun writing had, ifthe presented nimber was odd.

Anthmetic taskL The materiais and procedure wae the same as in acperiments fiom the

pnMaus chpter.

f the distracth tasks,

In both young anci older adults, the a n h c y chîndng urlr interfered substantiaily with free

d pedôrmance. In both groups, the odddigk tuL had a much malier &ect on manory. The

means for each madition are preserited m Table 7 for youag adults and Tabk 8 for older duits. The

bffowhg mdts wen signincmt at p < .ml unless othcmk nuteci.

Thae wae no significant main e&as or interactions with the orda -or (Order of

acpallneatal) on &ee recaIi monnance in the yaung duits. For the older aduits, there was no &kt

of Orda of arpaimeatal condition, but the Orda X Condition interaction was siBnificant @ = 3.16,

MSE = 2.61, p < .Ol). For one of the two otden of stperimntil conditions, in which the animacy DA

condition was the iast acpaimental condition to be paformeci, the mimba of words d e d was

disproportionateiy low. This couid accouat for the sipifiant interadion. Moreover, for h o s t di of

the participants (21/24), the mean number of words d e d was higher in odd-digit DA conditions than

in the animacy DA condition, regardiess of orda of +entai conditions. Separate adyses were

carried out ushg the numba of words d e d anci percentage âeciine in memory as dependent

TmiàEjres.

These a d y e s were d e d out using Ag* botq rid 016) as a between subject fictor and

E x p a b n d condition (fidi attention, DA liiimrcy, DA odd-digit) as a withm subject hdor. There

was no &kt ofAghg on mmiba ofwords recalled in the expaimaital coaditioas. Iinportantiy, there

wu no e f k t of Aging on the nurnber of worâs d e d m the nin attention @sselme) condition, g (46)

.- -- =h?R pr;@S, ~ ~ E w r s g U C C t s e f f i C i a ~ b i i s e l i a e ~ o f d piyamg&de.

n i a r r w r r a m r i n e f f c c t o f ~ c o ~ 0 1 1 ~ ( 2 , 9 2 ) = 3 0 . 4 2 , M S E = 3 . 0 1 . TheAgingX

condition intaraion wu non-signifïamt E (2,92) = 1.08, Q > .OS. The mean rnunba of

wods mcailed in the uillaas,, but not the odd-digit DA condition, M d sigdicantiy âom the mean

in the nin attention condition, E (1,46) = 40.39, MSE = 7.68. The Mèrence in numba of words

r d e d between the lallnacy and odddigit DA condition was signifiant, E (1.46) = 53.33, MSE =

4.39.

P- decline in rnemay under DA conditions,

For the foiiowing d y s e s I considered the mean pemntage dedine (ratio of perfiormance in

Full and DA conditions) for each participant as the dependent variable, Rtha than the percentage

decline of the meui absolute numbet of words recalied across these waditions. Analyses were carried

out using Agias (youq md old) as a betweai aibject W o r rad DA Condition (DA aaimacy, DA odd-

digit) as a witbm subject fkctor. There wcis m &act of Agiag on percentage decline in memory in each

of the DA conditions. There was a main &kt of DA Coadition E (1,46) = 57.79, MSE = 0.04, with

the DA animacy condition producing much mer dectims in memory thaa the DA odddigit condition.

The Ag@ X DA Condition interaction was signifiant E (1.46) = 6.63, MSE = .W. Q < .OS. S h p k

&ect~showstFktthepercentagedecfineinmemory

did not Mi for young and old duits in the animicy DA condition t (36) = -0.29, though it approached

Sgaincance in the odddigit DA condition t ( 36) = 1.76, p = -09. Surprisin&, the older adults

@ormeci better in the odd-digit DA condition than in the fbll attention condition, m g for the

. mteraction.

Expain#ot 4 - Numba of Words RecalIed, Auwrcy Elrtes and Reactioa T i (ii dikconds) on Distrachg Tdm, a d Paceatage Chaage From Swe to M t a s k Condition For Each Meenire for Youas adubJ W=24)

Measure and condition M S D

Target memoly task Words recaiied

FuII Attention 8.42 DA aPimacy 4.00 DA digit4 7.71

DistrPamg tasl< Acairacyme

BudineMimacy DA Mrmrcy Baseüne digit-id DA digit-id mer

Percentage deche in acairacy DA ammaCy DA digit-id

ELeoctioaw Iksdineanimacy DA anÙnacy B d e digit-id DA digit-id

Pemntage Uicnace in d o n t h e DA a h a c y 20.49 DA digit-id 18-34

Note: DA = divided attention

E q e r k n t 4 - Numba of Words Recaüed, Acanacy Rates a d R d o n Times Cui miüisecoads) on Di- Tasks, rad Pacentage Change Fmm Sm& to Duai-task Condition For Each M m for older dukr (N44)

Target memory ta& Worda rccaiied

Fuii Attention 7-17 DA .nimiCy 4-50 DA dia-id 7.2 1

DisCracting task AÇwscyme

Bs#linMimacy DA aaimrCy Baselw digit-id DA digit-id Filler

Badineanimecy DA animacy Besdiae digit-id DA digit-id

-

Note: DA = divided attention

F ~ d D A f o e d i t i o q t h ~ ~ a r a r g i \ o e p t h e E h a n r P ! t o r r a l l a i u r & ~ t h e

target tasic undar fidi attention. Few participants d e d any additional words. Young ad&s recaiied

ody 0.38 @Q= 0.65) wordr folowing the anhacy DA condition, and 0.25 = 0.44) words

fbiiowing the odddigit DA madition. Olda rdults d e d ody 0.92 @Q = 1.25) and 0.42 @Q =

0.78) words foUowîng the animPcy and oddaigit DA conditions respedvely.

Accumq rates (calculateci ris #Wl5 - # füse dumdl5) on the animacy and odd-digit

distracthg ta&, in the DA conditions, were much worse than siagie-task paforrnance. The percentage

dedme m acamcy rate was kger on the Mimrcy than odd-digit task for both groups. The meam for

each condition and age gra ip are prrseated in Table 7 md 8. There wae no sigdicant main as or

interactions with the order -ors (Order ofexpimental conditions and Orda ofsingietask measure

for each distracting ta&) on rcwacy rate, or perceritage deciine in urzuacy, for e i t k age group.

Separate Milyses were conduded using the accurPcy rate (hiw - Mse alonns) and percentage dedine in

acairacy (&om nagle-task) u dependent variab1es. - T 6 e s e ~ w a e c o o d u c t e a u s i a g A g i n g @ o ~ d o t ~ a s a b e t w e e n ~ b j ~ f f c t o r ~ d

Condition ( s b g b t a k Br a n b q and odddigit, ad DA for raimecy and odd-digit) as a wahin subject

&or. l'hem was no &èct of A@g on rcnaicy rate in esch condition F (1.46) = 2-80, p > -05. Thre

was a Mm &kt of Condition E (3,138) = 45.66, MSE = -01 1, but the Aghg X Condition i n i d o n

wts non-signifiant F (3,138) = 1.29, Q > .OS A#wscy nte on both dktrac@ tapLs uida DA

conditions diaatd sigmficliatly f b m thu mpedw siogle-task basdine pafonnanœ (F (1,46) =

These analyses werc h e d out udng Aghg (youag and old) as a between subject -or and

DA Coadition (DA animacy, DA odddigit) as a wahin subject -or. There was no e&d of Agbg on

pemmkge decline in icamcy in each of the DA conditions, E (1.46) = 2.70, p >.OS. There was a main

&kt of DA Condition E (1,46) = 19.35, MSE = 0.02, with the DA animocy condition producing much

farger declincs in accumcy t h the DA odd-di& condition. The Aging X DA Condition intaaction

was not signifiant E (1,46) = 0.02, > .OS.

An *sis was conducted ushg Agine boung and old) es a between subject fiutor lad

Condition (singietask for a n h q and odd-digit, d DA b r raimaçy ad odddigit) as a within subject

-or. nia+ wss no effcct of Aghg on RT in eech condition E (1,461 = 3.78, p > .OS. There was a

min e&ct of Condition E (3, 13 8) = 150.38, MSE = 8685.27, but the Aghg X Condition interaction

was non-signifiant E (3,138) = 1.29, Q > -05. Anothr rnriysis was conducted ushg Aging (jroung and

old) as a betwcen subject hctor arid DA Condition (DA rnimacy, DA oddigit) as a 4 t h subject

fâctor. T k e was wrr &kt of Aguig on percentage incnsse in RT unda each the DA condition, E

(1,46) =0.14, p >.OS. Thae was no main &ect ofDACondition F (1,46) = 0.39, and the Aghg XDA

Conditon interaction was not sienificant E (1,46) = 0.09, p > -05.

T h a r e a c t i o n t i m e s t o r e s p o n d o n a c h ~ ~ inthesingi~bsseImemdduPCtasL

conditions,rrrraDttdmT&le7d8. ThtbaseüneRTwirssignificoaUylongerfortheanimacythaa

odddi8it dktmthg task (22) = 9.53 for youug .duhs and t (14) = 8.20 for the olda adula) for both

age groups. The prœntage iiureare in RT, h m single to drukark conditions was not different for

t i m d m C y & * ~ * i & e i t h a ~ ~

The data h m ody 22 of the 24 young participants were aMihMe for this hie to

aror, data for 2 participants werc 10%

The rccuncy rate for both of the âbtmhg urks was smiüarly & id whai the suditory CRT

task was paformed col~curren@. There was a0 e&a of* otder on acairacy rates The mean

aca>racy rate on the uiimscy and odd-digit tasks, p e r f i o d concurrentiy with the CRT tom task, was

-69 m.= -14) and .71 (S.D.= -14) respectivdy. The ciiffierence between these two accumcy rates was

not signincant.

CRT tone ta&

The diffe~ence in the numba oftones comedy idenmed in each D A condition was si@cant.

The mean mimbar of correct responses for each condition is presented in TabIe 9. A wiuUn participant

ANOVA showed a main &bct of condition E (2,42) = 99.62, WSE = 40.94. The main &kct of Order

and the ûrder X Condition interaction was non4gnificant. Planned cornparisons showed the nwnber of

tones corredy idemifieci Un the animacy and odddigit DA conditions âifked significantly firom the fiül

attentionbudinecoadition(F(1,21)= 125, MSE= 116.49 andE(1,Zl) = 141.19, MSE=66.06,

rrspeçtivdy). The mimba of tones identifiecf in the smmPcy dual-task condition was significantiy Iowa

tbsn m the odddigit dud-task condition (r (20) = -3.03, p < -0 1).

Eqchent 4 - Numba of Comd Responses, Reaction Thes Çi MilliseCod) for Correct Responses on the Continuous Raction Tim Task for ELch Condition for young idults (N=24)

a 0 Condttion M M Baseline 97.82 22.41 793.90 177.19 D A man-made 72-09 17.49 956.26 174.0 1 D A digit-id 77.23 20.48 908.35 199.46

Note. DA = divided attention

The meui RT to i d e tones is shown for comct nsponses only (sa Table 9). An outlier anaiysi*s

eliminated RTs greater or leapa than two staadard deviations nom the mean for each participani in each

wndition. A within participant ANOVA indicated a main &kt of Condition F (2,42) = 16.5 1, MSE =

9271.5 1. There were no signifiant mrin e&as or intefactions with Order.

The mean RT in both of the dd-task conditions differed sienificantly from the mean in the

base& condition @ (1.21) = 31.83, MSE = 18218.64 and E (1,21) = 12.10, M E = 23821.35,

respctiveiy). The RT on the CRT ta&, pedionned under dual-task conditions with the rnimrcy and

odd-digit did not di&r (E (1.21) = 3.72, p > .05). These resuits iadicate ha î the two dküadng

truh make sïmüar resource demds or have simüar lmls of dîiEculty. 1 assume that the srune level of

diflticuity would appiy to the oldm duits.

Discussion

the DA conditions. Older idults, wah a pre~umed d u d o n in availabfe procashg resources, did mt

~ ~ ~ ~ ~ ~ H i - ~ t 4 ~ ~ ~ y o u a g ~ Tkcwofd-

based c iküdng task inttrfd substanMy with d e v d , but the sîze of the &kt wu not umplined

in the old comppred to yang ad& group. As in previour studies, DA at cetrieval using a non-verbal

distractiag tasic did not produce as iarge an efBct in either group. These findulgs support the

component-process mode4 account ofDA effects. As aiggested by F d e s and Moscovitch (2ûûû),

rêtrievd on memory tests mediateci p c i m d y by the MïIA proceed obtigatorily. It is oniy when there

is ampetition for m o r d systems, as h the DA condition ushg t word-based distracthg task,

that retrievd is disnipted. Because this system is b e l i d to be rdativeiy intact in older ad&, and

accordhg to the moded, nquires few genaal procesaing tesources for i ts operation, the pattern of

intedixencc e f f i is Siniü.r in older duhs to that of ywnger ad&.

If DA produceci interference by fedkcing the availab'i of g d resources, then older aduits

should have been Ifféaed even more than youiig when the word-bsed distracthg task was patormed

during reüievai. Moreova, they should Jso have had diffinity when the distrIctùig tssk was non-

vert& My r d t s suggest that the abüity to retrieve items h m wmory while attending to cornpethg

irrelevant material is not affected by agïng. The reduction in processing resources that characterizes

older adults does not Skct retnatal undex duai-task conditions.

1t should 6e notedthet the diff-ce ti Interference e f f ê h m the animacy compand to oâ&

digit tasic is wt due to ciiffierences in resource recpkements or ta& difficuity. The auditory CRT ta& in

thcyormsdutts,~odconarmPalywithecrchoftbtdistrPctingcsslcqdowedaduect~sSof

~tasicdif6aiityrridresourcerepuirrments. Resul tsshowedtht thetwo~tes l rshad

simüar~oaRThtbea1lditoryCRT~dthattbsricanacyratesonthedistnctmgEiskwat

notsi~~~iwlaDAcoQditioll~withthe~ae~etl~iny~adults~ This

! w g g m b t 8 i c b ~ ~ t n r t n ~ e q u i l l j r d i f f i r u t r r r d ~ œ d a n n n r i h p

In ime with pmrious work (Anderson et al, 1998; Mach & Buschice, 1983; Nybag et ai., 1997;

Park st A, 1989; WWng & Smith, 1997). DA ushg the -verbal c b a c h g trsk (odd-digit task) did

aot rne4 memory paformanct aubstantiiny. Thus, the prescrit resuits, as well as that of others argue

I&Pmst the hypothesis that reduced p d g rrsourcts or attentionai captcity in old aga accounts for

agc ciGmmes in episodic retrievai. Eit di4 dder ad& shouid have hd a harder thne cornpensating

fbt the reductiolt M availabie attention in e i k of our DA CO-.

It is p o m i however, thst in my sort to equate bosdioe levels of recall in young and oici, by

dowhg older addts to listen to the study üst twice iastead of once, 1 p r e v d the older addts fiom

exp«iencirig interférence which would othawise bave affecteci their mernory perfiormance. This may

bave given them more t h e to rehearse the target word li* and possibly to aeate mon robust memory

tracesthtarelesr~letodisniption, Thisporabitityisanimiaedinthcnextarperimant.

Uniike previous work (Anderson et al., 1998; Cnüc & McDowd, 1987; Whiting & Smith, 1997)

that a<amiaed paformance decrrments on the disasaiag crsk, the decline in performance h m singie-

to dual-task conditions for the odddigit and urllnacy ta& was not signincantiy elevated in older

compand to younger aduits, though there wu, a trend for higher costs in older adults. Fdure to find

b e r m g ta& CO- ih orda ddts wu ~arpeded Evai when acnuocy rate was m d

durhg ody tk fint M o f the DA condition, when most ofthe words are bàag r d & there was no

age diffkeaœ in distracting task costs.

Ifsuch costs d e e t the reso- aspect ofretnevai @e. rmintaniing mrieval mode,

monitoring output), ad are mediated by the PFC, then older cdults should h v e bem affected more

thsnourredt~mdicate~ PahepBthediff~hp~dwmgencoding.~eenyoimgPndold,

RT to d e 8 response to dhtmdng EpJk items wu longa in olda adults, though not

sigdicantiy. A h , the pacentage change in RT fkom Sade to dual-task conditions did not with

age. The longe RTs for the olda adula Wtdy occur due to diBimences in speed ofpfocessing

(Solthouse, 1985; 1996). Howmr, ramb rrnthscste timt thh difference in s p e d does aot lead tu

Merences in retrievai h m episodic mernory, once Uiitial leuning is cquoted. The mxt a<perimart was

conducted to d e the wonnurct of older adults under D 4 when the study opportwiity was the

same u that for yomg adults.

sipcriacicc

Introduction

Tbb arpaimcnt was conâucted to d e t e whether age-related dif€kences in H i of DA at

mtrievai would esnerge when iaitipl lePniing levds between older and younga addts were not equated.

It is possible tbat dowing older adults to listen to the study list twice p d u c e d a more robust mernory

trace that d e d pater susapt%iÜty with aging to disruption h m DA. In t h experiment I tested

rnother lgoup of oida amihs, anci allowed tban to mâytk arord Oa onty once, as the young amihs

did in ExpecVnait 4. 1 predicted tbat the pattern a d magnitude ofintaferetlce w d d nmPin the same

as that ofthe olda (and younger) aduhs in Eqerhmt 4. because interference fiom DA at retriatal

occurs due to cornpetition for coimnon repmentations, sid aot due to cornpetition foi g d

pfOceSSing resources, which are balieved to be c o m p r o d in older adults. By ümiting the number of

times the shidy list is heard to one, the overail numba of words d e d by older aduits in thh

acpaiment shouid be Iowa then thai of the older aduits in Expairnent 1. This prediction is in line with

the finding tbrt aging is associatexi with a decIine in epidc mmory (Craik & Jennings, 1992;

Sdthouse 199 11, that may stem h m diflicuities with enCodias md/or elaboratmion at encodiflg (Craik,

1982. 1983.1986).

Thu acpaimeat a b elIowed me to cornier waetlsa the rtruon I did not f i d the acpected age

diffierences in dhmcthg tasL mats, was tbaî I compensated for the oldet aduits' memory loss by

providing them aRtb additionai leMMg. Becsuse these COSU are Weveû to reflect the gerïemi

pmcesshg danrnQ of &ewd (Andetson et al., 1998; CnüE a al., 19%; Faneades & Moscovitch,

2000). oldardultqwboueprranaedtohwfewapmcsslmgfe~~~tcesavailabletotbem. shouid

hvebanmon~tbanyotmgaduhsbytbtDAcoQditi~a In~expehentIexpeçtedtha

addition to slower rrsponses to distncting ta& items, acctmcy rate shouid be affecteci to a greata

extent t h ~ was obiierved in the group of olda aduhs in 4.

Method

Partic@(~ntg

Partibcipaats were 16601der duits fiom the ENd.le Seniors subject pod. Each received

moadsry Compairation for puticipatiag. The man ige of participants was 69.40 = 3.09). AU

participants wae natm Eagiish speakers, and had n o d or corrected to nomial vision and hearing,

dative to thir age-matched peas. The Mini-Mental State ENOm (MMSE) was admhistend to each

participant, with those obtaining a score of Iess than or squrl to 26/30 king replaceci. The mean score

on the MMSE was 27.8 = 0.6)

Materials and Metbod

Materials end procedures were identical to EStperimait 4. The ody diffmnce was that the

study word iists for the acpaiiiLeatal session were piayed only once.

Resuits

As in Ekperhnt 1, the snllnacy chtmdng tactt pr&cecî a large interference H i on

mmory. whcraur the odd-digit distrPctiag tluL bd no effect on fhe recall pafomi~ace. The meuu for

each conditîon are presented in Table 10. The data wart MpiyEsd usiag &der of expimentai

condition as a between subject b r and Expairrmtrl condition (fidl attention, DA a u h q and DA

odddîgit) as a withai subject fàctor. The fo~owbg rraitts wue sisnificant it Q < .ûûl dess othawise

Thae waa an ~~ of Expaimsatll condition, F (2,20) = 9.84, MSE = 1.56, on fira recaii

perfbrmuice. niaewasaaa&aofOrda@(5,10)=3.80,M~=3.31,p~.05),wiUipoor~rd

in the last atpaimeat.l cotadition, but the Orda X Exp&med madition interaction was non-

sigdcant (E = 1.13, Q > .05). This indicates that & was gaiaally poorer in the last apaimentai

condition, regardles of whe&er it wur the ftll attention or one of the DA conditio~. The mean

wmba of w o d d e d when the animacy task was pdormcd collcu~raaiy with retrievai diffaed

siBnificantiy fiwn the mean in the fidi attention condition, E (1, 15) = 18.78, = 3.10. The mean

nudm of wor& d e d when the odbdigit ta& war the dbtracter did m>t diffa fiom the mean in the

fûii attention condition- The dinamce in m b e r o f words r@ed in the saVnacy and odddgit DA

condition was @ihnt, F (1, 15) = 7.52, MSE = 4.40, p < -05.

For the foIloWmg anaiyses I considemi the mari pemntage d& (ratio of pafonnance in

Full and DA conditions) for each partitcipant as the dependent variable, rather than the percentage

decüae ofthe mean absolute number of words recalled acrosa these conditionsIIs The mean percaitage

docliae in worb d e d differed for the two DA con&Oar @ (Tg = 258, p (0.05). Thae was a

iarger decline in the Smmocy tban odd-digit DA condition.

Expken t 5 - Numkr of Wor& Recrlled, Rates a d Reaction T i Cui dlbeconds) on Dutnctiqe Tas&s, and Paaatage Change From Single to Duai-task Condition For Each Me~surr for older ad& (N=16)

Tatget memory task Words recalled

Fuli Attention 5-06 DA 3.44 DA digit-id 4-88

Pacentage Dedine DA a h a c y 28-61 23 -72 DA digit-id -1.55 46-33

Dicitnctingtask Acaincy-

Bueiine animaGy -93 .O5 DA animacy -74 .17 Saseline digit-id -94 -06 DA digit-id .81 .18 Fier -89 . I l

Paazitege d&e in accuracy DA rnirmcy 20.17 19.90 DA digit-id 13 .24 20.94

Reaction time Btffane- M3-%? i 10.60 DA liiimacy 103 1.63 161.99 Badine digitid 624.13 121.72 DA digit-id 775.81 108.05

P ~ k m s e i n d o n t i m e DA lnimrcy 16-91 DA digit-id 19.22

Accumcy rates (calahteci as W 1 5 - # hlw rlirmdl5) on tbc .aimacy and oâd-digit

d k n t d q ta& in the DA conditions, wae d worse thm in the slligie-task condition The

pacentage decline m acniriey rate was larger on the ammpSr than odd-digit ta& though the diffuence

was not signifiant. The means for each condition are presented in Tabk 10. There were no signifiant

main eEects or htemdons with the order mors (Order ofacpaimentai conditions and Orda of

singIe-atadc meaaue for each distracthg task) on rccinacy nte, or pacentage dedine in 8ccuf8cy.

Seprate uiriyses wae conducteci ushg the rcwacy rate (b - Mse aiarms) and pefcentage deciine in

atnincy (from si@-task) as dcpadent variables.

khiwwmk

Thae was a mein S i of Expabmtd condition, E (3,45) = 10.40, MSE = 0.014. The mean

aamacy rites for each ta& in erch condition, are presented in Tabk 10. Acairacy me on both

distroaing tisTcs uada DA condinLons dBèefed sï@dkan@ coin theV respective shgle-task basehe

paformrnce (E (1, 15) = 18.01, = 0.02 and E (1, 15) = 6.43, MSE = .04, p < .OS, for the mimacy

a d odd-digit task nspectiveiy).

lblw€kr~L

Tha Ct(lCti011 time to rwpond on each d i i g ta& in the singie-task and duai-task condition,

is noted m TiMe 10. The singletask RT was Sgnificantly longer for the aaimacy thui odd-digit

distrachg task @ (14) = 8.20). The perceatage imesSe h RT, tiom siagie- to duai-task condition, did

not di&r for the two distrading tasks.

n of memorv Performance wrth . Ex~enmen t 4

Inordato c n n p s n t h e i n t ~ e f f ~ ~ t d u n â e r D A U t t h i s a p e r i m e n t tothatfoundin

the pmious one, the data tiom ExpaimenfS 4 and 5 w a e re-adyzed ushg an ANOVA with

E>rpauÜient as a between parîicipant &or. A separate analysis was carried out for the ?nimber of

wordp recaüed, pefcen&age décline in recd pafonnmce and percenbge &dine in Qccwacy on

d i a trrks unâer DA conditions.

The main &kct of Expiment on overaii lvunba of words recaleci was not signifiant.

However, pianneci Bonf in i cornparisons (with alpha = .OS) sbowed thpt oider duits in Expairnent 5

recalled fewa words in the fùll attention and DA odd-digit condition thon the older and younger ad&

nom Expairnent 4. The numba of words d e d in the DA a n h q condition was si@cady lower

for participants in E x p h m t 5 than young, but not older ad& fiom Expaunent 4. The &a of

expatneatrir Conditiion was 6B"ficant P (Z. 122) = 34.25, MSE = 267, though the Condition X

The main e&ct of E>tpamient on B e d e c h in d d e r DA conditions was not

sigaincariî, thcmgh the &kt of experimmtal Condition wasE(1.61) = 51.23, M E = 0.057.

Importantly, the Condition X Expairnent i n t d o n ans wt si@cant F (2,61) = 2.35, aad k e

~~~ Thdlecieeerqr----rtrrt8eurl,- iarger e&do on retrieval than the odd-digit distrading tasic, but this pattem did not differ depenciing on

age of participant, or number of sûdy trisls &ring encoding.

In the m d y i s of percentage decline in tcauacy on distncting tasks, the main &ect of

Expanaent was not siwcant, but the e f k t of- Coadition was, E (1.61) = 18.50, MSE

= 0.018. Notably, the Condition X Eqehmt interaction was not signifiant E (2, 61) = 0.47. There

waenodiffanrnasiuthepmntaged~h~ccuracy~ oneitherdi&actingtdq intheokkradults

Born Expehent 5 compand to those in Expriment 4 or compued to young aduits h m that

Discussion

This apaùnent was conducted to detamiae whether equating initial l e d g between older and

younger sdufts e h h a î e s sge ciifferences in effects of DA tht would otherwise be observed. Instead of

giving older aduits additional study trials, they teceivecl a single triai as did young addts in Experiment

4. As a dt, olda aduits in this expaimeat d e d fewa words than those in Experiment 4, but the

pattern and -de of interfice &kts remaïneci the same. T b , despite aiiowing older aduits to

Estai to the snidy word üst ody once d u h g encodiag, the inte&rence &kts fkom DA at retriewd

were not ampiifid in o l k .du&

As m the nnt experbmt, the word-bes#l d k h c h g task produced an approximaie 30"h

deciine m mernory nOm the fiJl maaion condition, whefegg the odd-digit distracting Essk haâ no e f f ~

on mawry. These rrsultJ nm counta to a reduced resource munt of interference effects h m DA at

tecrievaj. If the luge material-spdc &kt of DA at retneval ocnimd because tlns condition requirrd

~ ~ t o ~ ~ ~ ~ t b e ~ ~ s i h i l . , h h t u m ~ r e c o - v e r y a f ~ t h e n

olda ddts diould have s h m an amplifiecl Uitdbnce effèct.

In this cx@ment 1 iIso mvestiCgaied whether the absence of age ciifferences in diwscbing task

costs in E x p h m t 4 occurred because initiillesraiag wu eqpated between tha young and old. 1 had

arpected î q e r costs in older aduhs in Expehent 5. Despite worse memory, @orniance dectements

on the dbtmdng tasks uada DA coadition were not k g e r tha those observeci in the sample of older

ddtsinE><paimmt4. In~thepnscntsPmpleofddsrsdultsshowedmnsrmllacoststo

dWtraaing task perfiormance mder DA conditions with rdrievaL

Che might suggest that the reason for the lack of age clifferences in distracting task costs is that

our older aduits were mon high- fùnctioning than those used in otha studies. There are several

indications, however, t h tbis is not the case. In t<paiment 5, older adults showed the enpected Iowa

leveis of epiwdic memory chncteristic ofthe aghg popilition. They r d e d fma words in 1

conditions. Furthemore, the level of education of dder adults in our w p l e is aot dinacnt h m that

in otha studies (Anderson a al.. 1998; Craik & McDowd, 1987; Mach & Buscke. 1983; Whiting &

Smith, 1997). 1 Weve that the reeson I did not hi greata distracting task costs in our olda aduits is

that our distncting tasks are not seasitive enough to d i f f i e the age groups. The sensitivity of my

caosen&sfncbagt.sttqtodisaimiaat a . . e tevetsofakresoarces, isexaniindirttkncxt

E.perinent 6

Introduction

in Eqmhmb 4 and 5 1 used a populrtioa of olda ad* to test whether cornpetiton for

p n e r a i p r o c e u i a e ~ c a i l d ~ f ~ r D A e f f i p d u c e d d u r h g r r t r i e v a L Themuonfor

testing olda adub was tht a&g is pmmmed to be rarori.ted with d&cits in memory stemming h m

a reduction in generai processing resources that am &cal for numemus cognitive operations (C&

1983; Craiîc & Byrd, 1982; Rabinowitz, C e & AcLamui. 1982). These resoufces have been

conceptunüzed in many ways, though most theurists suggest that changes in fiontai lobe ninetion

undalie the deficits (see htmduction to this ctiripter for d d ) .

Anotha brria region shown to decline wah age. tbtt b dso relateci to episodic manory, is the

medid temporal lobes (Moscovitch, & Wiwan, 1992). The MïLlH has long bcm considaed to be

cniciai for episodic memory (see Shimunna Bt SqUm, 1987 for a review). The cornponent-process

model, wbidi d e s c r i i the role of tbese different bmh regions on di&rent tests ofmemory, proposes

tbat the MTL/H is a moduiar system, and thus does wt coatniute to the h e i ofDA at retrid. The

r d t s of my previous work are consistent with tbis claim (Fernandes & MoscoVach, 2000, Fernandes

& Mostovitch, submitted). In thU exphent 1 use pediorrnanœ on neun,psycho1ogiical tests, as an

adanal merairr oflevd of W o n M g of the âoatai and EnLZIR, in a group of olda adults, to

d u a t e more dirsctly the contn'bution of thse regions to interference &ects nom DA at rrtrievaf.

Thae is evidence that sugeests thot aot 1 cognitive fuuctions are simüsrly &ècted by agi118

rad the rate ofdeciîne may diaes dependiag on the &en Gaction. Individuals also may diffa with

respect to wbich M o n s are most rffccud (An#rs 1988; Welfbrd, 1993). This wiobility moy have

praventedmfrom~sneff~~tofAgmeoamt~inthe-ryooddistractingt.sLs.

- - -, E % k e f d R a i t h i e r r e t C & 9 9 5 ) t o d t ~ & ~ ~ @ ~ ~ g ~ o u p e

of olda aduhq dependhg on thek levei o f ~ o n h g . Ushg neuropsycho1ogicai tedng, the Id of

âontal and temporel lobe M o n for cach iadividud was meseci, reiativt to the group as a whole.

They conducted a factor adysis on a seIection of neuropsychoIogiicPI tests thought to depend on

hntai and temporai lobe M o n . From thk, they were Iibk to dculate two z scores for each

puticipant, one tht represented a composite of âontrl lobe (PL) W o n , and one representiug

f Û n c t i o ~ o f t h e ~ t a m p o d ~ ) l o b e r . Thywerethnrabieto~thürpoolofolder

aduhs iecording to ievei of fiontal d temporal M o n (pee Güaky et ai., 1995 for details).

In the presesû experhent, I considemi the paformoace of olda addts in Gir*sky et aL's

popuiation, unda the some DA conditions as the participants in Expaiments 4 and 5. Gimi that my

previous work (Fwmmdes & Momntci~, 2000, submitted) mggestecl compaition for EITIZ. did mt

rmdcrüt interfietence effect8 under DA conditions at reüieval, 1 predict that the Id of temporai lobe

action s h d d not Jter the magnitude or pattern of i n t e r f i i effecf~ Etom the ociddigit and animeey

disacting tasks. Furthemore, the e&ds of DA rn bdmd to be due to cornpetition for

npmentations in the posterior neocortex, thris wkther a participant has a high or iow Id oftemporai

lobe M o n sbouid aot influence the interference Ushg the same reasoning, the kvel of FL

niaciionhg M d mt chan~e the magdudë or pattern ofeffects ofDA on memory. Ekperhents 4

and 5 lnowed me to infa that FL fimctioning does mt c o a t n i to the interfenace dect on memory.

Intbepment ~ I ~ t h a t d o f t h e p r r t i c i p a i i t s w i l l s b o w a s i m ü a r p i t t a n a a d m e g m n i d e

of memory mtarftrrace to the olda ad& in E.irpaiment 4 ind 5. Because the MTL is important for

episodic XneYnOry, Qiring both eacodmg cmd retrievd, howeveq those with lowa level!? offiioctoning

sbould recaiî ficirnrwotdsundafiill attention caiditioas thsathosewithnormal h d s offuadoning.

hsBq9ekmw4seO5, E B d ~ * - w o k o a F ~ i a t a s a c a w ~ o < r t b c

disûadq tuLs, as dis muure is believed to dect the resource-demanding mpect of retrieval

(Andason et ai., 1998; Craik et pl., 1996; Fernandes & Moscovitch, 2000). Establishing and

. . . mnintrrmnp retrîevai mode, rad monitoring output, are rll bdmd to be meduted by the PFC, wbich U

p d l y compromiseci in oider adults- The iack of rn rmpiified &éct of DA on tesk

~atscouldkiatapretedtomtuiththsampkofo~k~inEitpaiments4oisd5 didnot hawa

reduction in g d process@ rmaurar, and arc llOt nprrsentri3ive of older addts in the grnaal

popul.tion. Alternatively, it cotd mean that costs to our chosen distractin8 tesb do aot rdez t the

resourcedemands of retrjeval. The prrient apaimeat dows me to distinguish b e e n these

aitanrtive interpretations.

1 w m p the Mormance of older dults with hi* versus lowa levels of fhntal lobe

fÙdoning, retative to the grwp as a wlsde. Baause 1 wül have an indepadmt measure of fiontai

lobe Wonin8 with wbich to chssî& olda addts, 1 can conàder whether low levels of M o n lead

to higher distrading task costs; this is expected ifcosts d e c t the resource-demanding aspect of

retrievaî, which are prernuiaed to k mediaîed by the hnsal lobes.

Because gMag olda adults additional shidy trials did not aiter the magnitude or pattern of

&&ts of DA we ci(lowed the grwp of oîda putia'prats m taiP sais to liaen fo the shrdy list twice-

PartichlQ

Senion were sdedsd h m a Iarger pool of herhhy, cormDuMtydw&g duits over the age of

65, u b undergone adeawre naaopsycûoiogical teshg wahin two ycan of e x p h m t d testtig at

theUnmratyofArizo118. ~psrticipmtinthepoolbad beenllssignedtwo scores, onenpmsenhg

~ ~ o a & ~ o f ~ ~ & E L ~ u Y i t b e o t b c r ~ r r l r t i u e

paformancc on a p u p of tests bdieved to a s i ~ a ~ kTCZ fiiocton.

The tests comprishg the FL fàctor wae: mtmber of categories achieved on the modifieci

WlsconBa Cud Sorthg Test @art, Kwamis, W d e & Taylor. 1988), the total number of wor&

generated on the F M test (Spran & Benton, 1977). MentaL Arithmetic fiom the Wechsier Adult

Intelligence Sale - Rm*sed (WAIS-R; Weciider, 1981), Mental Control âom the Wechsler Memory

Scrk - M (WM-R; W&, 1987). d Bwkward Digit S p EOIII the WMS-R

The tests comprishg the MTL f a o r Mude: Logid Memory I, Vabal Psired Associate I, and

V i P W k o c b t e II (ali h m the WMS-R) anci the Long-Dday Cued R d m m from the

california Verbol Leamllig Test (CVLT; Ws, KrPmer, KnpiUi, & Oba, 1987).

These tests were grouped 8ccording to the d t s of two fàctor analyses: a) an explontory

principle fkctors analysh of dota from 48 older adults (reportecl in Glisky a al., 1995), and b) a

confumatory -or d y s i s of data h m a separate gmup of 100 older adults. In order to assess the

differential coatri'butions of the FL and MIT that were mdcpadent of age. miance attributable to age

was removed h m test scores pnor to analyses. Composite scons were caldated for each individuai

in the pool; these represait average z-scons for those tests loadùig on each &or, relative to the 100-

memba n o d e group. Scores in the 100-~t~1nnativegrorip were d i s t n u spchdtat34 are

above the mean for both the FL and MTL tiictor (denoteci AH group in this apallnent), 26 were below

the m e ~ on both Mors (denoteci LL group in tbis e>rpamieiit), 18 wen a b m the mean on the FL

f iaor and Mow the mepn on the MTL âtdor (deaoted HL grwp in this -1% and 22 were

abovethemaaonthe~hdoraiwiMowtbemeenontbe~~r(dmotedLBgmupmtlns

eirpallnentl*

~ 0 ~ t s O ~ ~ 2 2 p 1 F t i c i p I R t c ~ s d A C l t P A ~ t k a l p o o i 0 f 1 0 Q * d t r ; o

b m the HH group, 5 h m the LL group, 6 h m the HL group, ad 5 from the LH group. niey

ncaved n m e b q comp«lution for th& participation. Charadétistics of each group are pnsaited in

Table 11. Separate one-way between-subjccts ANOVAs iiadicsted that there w a e no Merences in age

oreducation as afuactonof(jroup. Therewrsui&ectofGmup onMMSE score @(3,18) = 5.71, p

< .OS); a Boderroni pst-hoc cornparison sbowed thst the LL group s c u d Iowa than the HH and LH

pup, howcvsr, scores grma thui 26/30 an considend indicative of n o r d fùnctioning. There was

also an &kst of Group on fiill scak IQ (E (3,18) = 4.14, p < -05); a Bonferroni pst-hoc test showed

this & i was due to the U grmp baving a much lower IQ thDn the HH group.

Cbaptcr 3: DA in young and old ldultr Table 11

Mean Chatactenstics (M + SD) of Groups S e l d Accordhg to Frontal Lobe @) Function and Medial Temporal Lobs (MTL) F d o n md Extirmt<#l of Published Normative Data on Neuropsychological Tests for Aduîts Ag4 65 to 75 Yous.

Wisconsin Card Sorting Total words on FAS WAIS -R

Mental Arithmetic WMS -R

Mental Controi Backward Digit Span Logicai Memory I V a W Paired-Associates 1 Visual Paired-Associates II

CVLT Long Delay Cued kcaü

Composite Score FL 0.7 + 0.4 -0.6 + 0.4 - 0.5 + 0.4 0.7 + 0,4 - MTL 0.7 + 0.4 -0.5 + 0.3 0.5 + 0.3 -0.4 + 0.4 -

HH = high fiontavhigh temporal, LL = low fiontsvlow temporal, LH = low fiontai high temporal, HL = high fiontal, low temporal

M.terials aad procedures were identical to those of-t 4. only

Mereace in procedure was that participants in this acparimnt were given only a short

break in between the experimental session

Results

Tarset memov task mdy& In ail groups, the animacy distncting task interfered

sub- with fhe r d pafonnmcce. In aI groups, the odddigit task had a much

d e r effect on memory. The means for each condition are presented in Table 12. The

following results were significant at p < .001 unless otherwise noted.

The e&ct of the Order actor was oon-significant as was the Condition X Order

interaction. S e e analyses were conducted wing the number of words r d e d and

pacentage decline in memory as dependent variables.

of words recalled,

These analyses wem carried out using Group as a between subject fanor and

condition (M attention, DA rnimaCy, DA odddigit) as a within subject

fhctor. Thae was no &ect of Group on number of words d e d across all experimental

conditions. Thae was a main &kct of condition E (2,36) = 1 6 . Y =

2.92, but the Group X Expaimensal condition hnta'.don was non-significant P (6.36) =

0.46.

Aaoaodgmupqthemrmbaof~~rdsdedintbeDAlloirmcyconditionwes

SignEcady less tban that in the firll attention COQdition, E (1,18) = 27.63, MSE = 6.60.

The nrrmba of words recalled in the DA odddigit codition, howeves, did not o t d i f f e r h m

that rmda a t t w E (1,18) =4-09. The diffZrence in mrmkr ofwords d e d

= 6.93.

A DUNCAN multiple-range test @ = .OS) shows that the HH lpoup recaüed

of wods recaiied uada fidi attention, depending on frontai lobe c l d d o n (iiespeaive

o f MIL grouping), E (1,ZO) = 0.47. Then war a signifiant Metence in words r d e d

uada fidi attention, depading on levei of temporal lobe fùnctioning (iiespsctive of

fiontal lobe grouping), E (420) = 7.60, Jkl@ = 5.83, Q < .W. Adults with low MTL

fiiaction d e d féwer words thn those with high MTL hction.

Across groups, the percenta%e decline h memory, under DA conditions with the

aaimocy tasic wk( siflcant, t (20) = 5.91. The decliae under DA with the odd-digit task

wu non-sigdbmt, t (20) = 1.01, p > O S .

The raniining mdyses were conduded using Group as a between subject h o r

anci DA Condition @A animacy, DA odbdigit) as a wahin abject fkctor. There was no

S i ofGroup on the magnitude ofdedine m memory <mda DA conditions (F (3,lS) =

1.11). There was a ~MÜI e&ct ofDA ConditionE (1,18) = 17.22, MSE = 0.057, with the

DA mbacy condition produchg much kgex declinCs in memory than the DA odddigit

mdition. TheGraip XDACandit ion~onwasnot~Bii i f i~811f~(3,18) =0.86.

T h e r e w e r e a o d i f f ~ m p a ~ e n t a g e ~ i a ~ r y , d ~ g o n l e v d o f

~ 3 t M i m ~ u d d d ~ 104

fiontai Iobe M o n , F (1,20) = 0.35, ad F (1,20) = 2.09, unda DA conditions with the

uiirmcy rnd odchiigit task nspectinly. Thae were dso no no depending on

lsvd of taapod loba bction, F (1,20) = 0.08, and F (1,20) = 0.95, unda DA conditions

with the rnmucy rnd odd-digit tkck reqe&eiy*

Expsriment 6: N u m h of Words Rccalled aqd Percuntage Change From Singie to Dual-task Condition For Ewh M-e in Eich Group

Memure and Gmup Full ~~ent ion DA a h a c y DA digit-id M SD M SD M SD

Words Recalled

E o l t o a m 3 g r r h n l ~ t h ~ œ r anrr*eQthechrncetQrecaIL

worcb h m the target task under fidi attention. The nimba of additional words recaîied

following each DA condition, for each group, are presented in Tabie 13. A DUNCAN

dtipie-nnge test @ = .05) shows tbat the number of dditiod wordr d e d followiag

the DA anbacy condition did not Mer acroar groupq the numba of additional words

d e d fi,Uowbg the DA odddigit condition was higher for the p u p with low

fiocttamgh tempod fllnetom

Becouse the number of Pprticipants in each group was smrll, data was coihpsed

aaoss 1 4 ofFL and MTL f ù d o r ~ Therc was no sigaificaat diffaence in the n u m k

of additional words d e d following the DA enimpCy condition dependhg on frontal

lobe ciasMcaiion (imespdve of MiL gmuping), F (1.20) = 3.97 or depending on MTL

damification Çleqedve of frontal lobe grouping), E (1,20) = 0.06. Similady, thae was

no siemficant differerice in the numba of additional words r d e d foiiowing the DA odd-

digit condition, dependhg on hntai lobe c l d a t i o n Cmespeaive of MTL grouping), P

(1,20) = 0.53 or dependhg on Erîn clRQQification C i e of fiontal lobe grouping),

E (1,20) = 0.39.

Experimais 6: Additional Words Recaned FoIlowing each Divided Attention Condition for -Group

HH = high frontaVhigh temporai, LL = low fbn&ylow temporal, LH = low h n t d high temporal, HL = high frontaI, low temporal

Acaincy rates (calculatecl as #hW15 - # fàlse rlsnnsll5) on the ruiimncy ad odd-

di& distrachg ta& in the DA conditions, were much woot thn unda singietask

@ormance. The percentage dedim in rcainçy rate did not diffa in the DA Ugmrcy

and DA ociddigit conditions, amss groups. The meciar for each condition and age lgoup

are p d in Table 14. Thae wae no SiSmfiant mUn &écts or interactions with the

orda fàcton (Orda ofacpaimenîaI conditions rad Orda of ynsletask meaave for each

distrachg ta&) on scavacy rate, or perceutage deciine in acamcy. Separate analyses

were conducted using the ~ccuricy rate (hits - hlse dams) ad percentage decline in

~iccurpcy (hm singîe-tasic) as dependent variables.

Chapîcc 3: DA in yourg and dd adui8a

Table 14

Expriment 6: Acwroy Rates on Distrqcting Twk~ ud Percentage Change From Sigle to Duai-taak Condition For Each Mcasum iq I3rch Group

Messure and Condition Bwiine animacy DA animacy Baseline digit-id DA digit-id '

M SD M SD M SD M SD

Percentage decline in accufacy

Note: DA = divided attention

Tabla 15 Expiment 6: M o n Times (in miuisewnds) on Distracthg TiisLa and Percentage Change From Sigle to Duai-tssk Condition FOE Each MealMe in rnh Group

M a u r e d Condition BpJsiuieanimirCy DA a n i x ~ ~ ~ BPsdine digit-id DA digitid ' Ad sD M SD M SD M SD

1

Percentage incce~~8 in reaction time

Note: DA = divided attention P

C h p t e r 3 : D A h y o m g a n d d d ~ - There d y s e s were conducteci ushg Giwip L( a betwcai subject -or and

Cobditi011 (singk-task for aaimacy ad odddigit, and DA for cinimyr a d odd-digit) as a

within subject fàctor. There was no &kt of Ciroup on acumcy rate in esch condition F

(3.18) = 0.83. Thae was a main &ect ofCodition E (3.54) = 13.50, MSE = 0.008, but

the Group X Condition i n t d o n was non-significant E (9.54) = 0.73. Acatracy rate on

both distroctiqg tasks unda DA conditions d i f f d sigdicently nom th& respective

single-task brsdine pafomiract @ (1,18) = 21.33, MSE = 0.019 and F (1,18) = 26.61,

MSE = 0.010.. for the Mimacy and odd-digit tasic mqectively).

Across groupq the percenfage deciine in aCCUCBCy rate frwi single to dupl-task

conditions, on the aaimocy and odd-digit was sipiûcant, (20) = 5.06, and t (20) =

4.77, rrspectnrely. The= were no Merences in percentage d e c h e in accmcy, dependkg

on Id of firontal lobe nuiction, F (420) = 0.04, aad F (1.20) = 0.54, for the animaçy and

odd-digit ta& rcspectively. There wae dso no dinamces. depencüng on level of

temporai lobe hction, F (1.20) = 0.94, and F (1.20) = 1-77, for the animacy and odd-

digiîtaskrespectntdyY

The renieining dyses wae conduded using Group as a between subject fictor

and DA Condition (DA a n i w , DA odbdigit) as a wahin subject fsdor. There was no

eifiéct of Grwp on pacentege declim in acamcy in the DA conditions, E (3,18) = 0.35.

Tben was no mrin &ect ofDA Condition F (1,18) = 0.52, and tbe Grwp X DA

Condition interdon was mt Sgdicant E (3,l8) = 1.41.

sepamte ANOVAs were coaducted usbg percentage decline for each DA

~ * P & & ~ ~ 0 1 4 ~ Il2

condition (iindMduiUy) as the depedent variable .ad Group as the between subject

fador. Thaew~noeffectofGnwponthepacmtaeedediaemmanoryinUieanllnricy

or odddigit DA conditions, E (3,12 1 = 0.3 1 and E (3.2 1) = 1 -85, respedvely.

Reaction T i k

The &on times to respond on esch dhtmchg tpsk. h the single-task badine

and duai-task conditions, are noted in Tabk I S. There was no e&d of order of single-

urk wuwos on RT, for either of the distnctllig tssLs in MY group.

The pacmteee increase in RT on the distractin8 tasks unâer DA conditions, did

not differ dependmg on levd of âontai lobe function, F (1 JO) = 0.21, and F (1,20) = 0.11,

for the anhacy and odd-digit tasic respeçtiveiy. The imxerse in RT did not Mer

dcpardiag on levd of temporai lobe fùnctiou either, F (1,20) = 1.49, and F (1,20) = 2.27,

for the mÛnacy ad odddigit tisk respectiveiy.

The followhg analyses waa conducteci using Group as a between subject fector

and DA Condition @A Mimacy, DA odd-digit) as a withlli subject Wor. Thae was no

&kt of Grwp on RTs in the DA conditions, E (3,18) = 0.53. The &kt of DA condition

was si@cant, E (1.18) = 41.37, MSE = 68S4.15, but the Group X DA condition

interaction was non-signifiant, F (3,15) = 0.06.

There was no &ect ofûroup on the percentrge maesSe in RT on the distnctllig

tksics unda DA conditons, F (3,18) = 0.25. The &kt ofDA condition on the increase in

RT was m > n - s i @ ~ as was the intaaction, F (1.18) = 0.21, and F (3,18) = 1.02. The

~mRTwarsiniitrrforthernmiocyrndodddi~dismctiagtps4F(1,18)=0.21.

I useci duhs clmsifled on the bais of neuropsydmIogicd test scores. as having

good or poor FL and MTL fiiacton, to d u a t e mon dllealy the contribution of these

regions to interfefezlce &eds âom DA at retnevai. Accordhg to the component-ptocess

model, the dkts ofDA a r k from cornpetition for rqmmtations, in the posterior

neocortex. Thus, whetha a participant har a high or low l d of temporai loôe or fiontai

lobe M o n shouid not influence the extent of

The fiadins tht levei of meciid temporai lobe fùnctioo does not contn'bute to

interfefe~ce d e r DA, is consistent with my pmious wodc Fernandes & Moscovitch,

2000), m which participrats had to recaü words and monitor simuitaneoudy a iist of

VisuaUy-presented words in a running recognition test: in one DA condition, some words

wererepeatedafkatttlagof3 orless, andmthcotba,atalagof7ormore, makingita

test of short-tem or long-tam memory, rcspedRreiy (see TuIving & Colotla, 1 970). Both

DA conditions wae succedbi h producing a substantid decnment in recd of about

3% in cornparison to the fidl attention condition. That there was no difkretlce between

the short and long hg conditions suggested that INIL, which is involveci in LTM but not

S m was not the locus of the competition e f k t betwear the met and interfering

memorytaskstasks Instead,tbtdtssuggestedthtthet~~~~ofthteffectwasetthlevd

of neaeorticai representatiod systemr hoIved m word perception and production.

Furthennoce, theresubare~liaewiththosetiomExpaimans4and5, mwhich

the dudion in avaiIabIe p&g nsourîes believed to cbaracterize dder ad* and

atkiibuted to deterioration in the hd lobq did not dter the pattern or m a g h d e of

mtdbmœ. If cornpetition for resources accouats for the lerge &eds of DA âorn word-

C l u p t c + 3 s D A I r y ~ 8 u l ~ ~ Il4

baseci digbacting tpsks. memory i n t e r f i in participants with tdativeiy poor FL

fimction shwld have beai smpüfied rdatiw to those with better FL W o n , bscause the

former have to compensate for an even greater reductioa in rvailaMe rrsources. Thus the

r e a b Bwn this experheat argue ie;iinst a nsourcc iccount of DA effects at retrieval.

The ody effiècts on memory were seen on the nimba of wordr mcaiied under fidl

attention: those with reiatively poor MTL b a i o n d e d sigrtüicantly fewer words.

This u m expeded bcuuse the MTL is important for epoodic memory. Regardfess of

thek lowa brrdine m e m ~ r y paformrncc, the pattern and size of tfieir interférence &ect

was no different than thai of the participants who had rdotveiy higher levels of MïL

function, Thus 1 have shown that whelha the o v d Ievei of memory perfiomrence is

affecteci natudy, in those with poor MTL fîinction, or artificiaüy by increasing the

number of study trials. in the h i ~ ~ o n i n g olda odults âom Expriment 4 and 5, this

does mt iduence the magnitude of mcmory i n t e r f i h m DA

As in Expaimaits 4 and 5, DA Usag the aaimrcy tasic produced v a y large lmls

ofinterference, ran@ng from 2 7 4 % drop in perfionnance h m fU attention coaditions,

m s s groupa Smaila interfice e f k t s were observeci when the odd-digit talc was

paformed with d (range -23% to 22%). These data support the clah tht effécts

fiomDArtntrieVasremodulatedbythetypeofmat~ inthtdistraamgtask. Thus,

data m this cbapter show that matnial-specieüty of the distrectiiig task is a major &or

determining the magni-aide of the e&a on memry.

InEqehmtc34md S,tbekcLofmamplified&~ofDAondisbadingtssk

coariswrsmexpxtd. TbiscouldkiaterpretedtomaathattheYmpleofolda~in

theseerrpamiansdidwtbavearrduaion~genail~resources,andwaenot

£ I l # p ( e r 3 : P A & ~ d d 4 ~ US

fep- of& addts in the garaol popdation. Alternativelys it couid mean that

coststoourchoaendistm&gtasics& aotrcflectthe- ofretrieval.

Thisaq#imentallowedmetoaddresstbeseiSsues. HaeIhadanUadepeDd~~~fmwure

of fkontal lobe fmctionïng with wbich to dLIsify older addts, end could consida whaha

low lcwb offiiactoning l d to higéer diJtradine task cos& this is expected ifcosts

n&a the resource-demading aspect of d e v a l , mcdinted by the &ontal lobes. Thae

was, huwever, no d i f f ~ a r e in costs to eitha disüa&g regardIess of whether

accmaqorRTwiuthemeasurewed, betweengroupwit6reiativelygood orpoorFL

on.

Nevertheiess, the finding ofsigniiicant costs for each task, unda dual-ta&

conditions, suggests that retrieval resourct-deimanding. W h t is more, the costs to the

distraaing usk did not Mer for the animacy and &-digit tssk Thus unlike the effects

ofdinaent diPtricting ta&s on memory, which are matefial-speafic, &écts on the

distrrctmg task naut reflect some more g d resource repuirement.

G e n d Discrusion of CLipter 3 Ehpdments

The p d chppta was aimed at detamiriiag wbahalrrge interference effects

b r n DA at retrievai d d be accounted for àtber by cornpetition for gened processiDg

n#nrrces or for common r e p r d o n d structures. The work by Fernandes and

Moscoviîch (2000), as wdl as the work d e ~ ~ l i in Chapter 5 suggests tht @ormin8 a

word-based comammtly with d of words, mtafnes with reactivEUion of the

neocortical repmmtations that form part ofa vabrl mmory trace. Cornpetition is

aerieduthemmorysaddWtnctingtksLswmpdctOrthesrmnpr~omalsystem

€ b p 4 e ~ & B A t y e o r g & s l b m U6

An~.wwntoftheseresuitsiPthttheiarge~rywsts,fiomsinnla

mrtairI in a diarsctiag tu4 are due to a rechlcfion in available resources for f?ontai lobe

fbdon.. Tbe# nswcer, may be isaded to coordllrrte tbe oaline procaSeng of durl-

tultr. In the case of the word-based dhtndng tuLq ntnevsl miy be disnipted to a

re ia t ive iygreateracta i t~ t&es imi lMtyi l lmrta ic i l s~ i tmore~al l t to

adnct dmnt items h m memory, d o r cootdinate thcir output, both ofwhich

overextend a limited pool of gaaal proceshg resoums. To compare these accounts, 1

c o a i i d d the paformance of ofder aduhs, whose cognitive tesources are believed to be

reduced, W v e to young adults.

kbnmQm

In Experhents 4 and 5, youag Md old duits wae sllnüuly affiected by the DA

wnditioos. In both age groupq mmory was mch more af'fécted by the animacy

compareci to oâd-digit distractin& tsrlc Importantiy, olda aduits did not show

interference &eds on memory t h were hrger in magnitude to those observeci in young

adub (see Figure 3).

. Young

. a d Expt, 4

T 001dExpt 5

animacy

Distracting Task

Eigurr 3. Mean pacentage decline in âee d paformance fiom fidi to divided attention @A) for ach ColKiifion m E x p b e n î m.) 4 and 5, in yomg and oider aduhs.

bahpmiouashiclies(Andaranct s1,1998~Mh&&Buschke, 1983;N-et

A, 1997; Puk et d, 1989; Whithg & Smith, 1997) DA at reûieval with a not~verbal

distncinigtrrkdK1wtpducevay~eeffed~onmaaory,dthaewasno

interaction with age.

These resuits argue agabt a duceci resource exphdon of interference &éas

on manory, and support the idea thit retrievaI is disrupted primarily when there is

compdtbn for c o n npresentatiod syûems. Moreover, ifcampetition for fesources

accouats for the hrge effécts of DA fiom word-based distracthg tas& memory

interference in participants with reiativdy poor FI, nincton in Experhent 6 should have

been unpüfied relative to those with reiativeiy better FL ninction, but it was not (see

Figure 4).

Distracting Task

Figure 4. Mean percentage dedine in fie+ recall h m smgle to dMded attention for older ad& with r e h j d y high versus low hd lobe W o n with each distncting ta!?&.

a d t h p a f o d Ma in the odddgit DA conditions than unda fidi attention during

retrîevai. 1 have no expianation for this finding, except tht p a h p s a d h g a simple

distnctmg tssk raises arousd, which improves perfibmance (see Kahneman, 1973).

An aitefiaative interpretation to îhe composent-ptocess d e l , for the large

materid-spccific DA &ect, is that compatition W cnated at the Id of inputsutput

channels in working manory. As suggested by Hrs&r anci Zacks (1988), age-relateci

diffaaices in memory uid otha cogdke ~ ~ & O L I S am be attnhrted to a dediDe in

attentional inhiiitory control over the contents of working mernory. They suggest that

older duits are more distreaed by irrelevsnt information: reduced inhibitory control

allows more 'hongoal" i n f o d o n to enter working memory, thereby produchg diflCiCUfty

with memory. If DA at retrieval in my atpaimsnts digupted memory by this means,

however, olda adults in Experllnent 4 wodd have abown greater interference in the DA

anmiicy (and DA odddigit) condition rdrtive to the ywng adults.

One priLtüng Mèrence in the findiags acrosir experllnents, however, suggests that

cornpetition for output paShwrys m y be o t h e r source of interference under dual-task

conditions. Ln Exphmts 4 and 5, both young rnd olda participants d e d few, ifif,

additional words &a the DA condition was tenninated That is, the &ectS of DA h m

the word-based dùtractiag task pasisted in both age groups. Such results are consistent

with the notion tbrt word-based dUtnctmg trsb iilterfae with ~ m o r y by colfll~tmg the

~rytnce~~es&M0~~0vitcb,2000~priortoaSbQagrrport~d

inmasistent with the possiiilitg of competition fbr mput-output chronels in working

memory. In 6, ho-, only the E H gnwip showed the srm paSartent

~ ~ D A l r y o r r i r g u d d d l r r l p j h 121

eîEcts ofDk The otha grwps of olda duits d e d many more words a f f a the DA

COaditi011 WûS QVW, t0 the p0a 0ffecovef@ tbe m - S @ f k Ülterferen~e

e n é c t f h m t h c a n i m a c y ~ t a s k Theae~finAingsmmoreùiiinewithan

output inrafaace account of DA e&cts. It may be that participants with poar MTL

d o r FL W o n f d the dd-task conditions so difficuh that they did not try to recaü

words whüe doing the dküadhg tulc That is, tby rmy bave switched baclc and forth

~ e e n t b e t w o ~ ~ t h o a ~ t o c a n y o u t b o t h s m P i l t a n e o u s y e s d i d t b e a o n n a l

fiinctioxhg dderly (Hl3 group), and the older aduIts f b m E x p h m t s 4 and 5. This

suggests that memory traces are not corrupted unda DA conditions unless one is activeiy

tryiag to wry out both tasks simuf-, rather thPn in tandem. Despite this anornaly

Li additional worâs d e d foiiowing the DA conditions, there was stil i a larger

interfetence &kt h m the raimacy compand to odd-digit dUtncting ta& Moremer, its

msgintude was d a r regardlets of levei of MïL or FL bctioa Thus cornpetition for

outpit chpmids is not the onfy source ofbterfefetlcf.

Distractinn Task Costg

Accordhg to the component~process nmdei d e s c r i i in the Introduction, the

fes0urCbd-g aspect of ntrievai lies Ui estabürhiag and mahîahhg rrtrievai mode,

as weü as m monitoring output. These pmwses are t&ought to be mediritecl by the PFC.

lasofiu as the target task mrkw use of that proceses. the resource danaads s h d d be

refkacd in cortr to d h a d q tasic p a f o ~ . Tndeed nich COSU were rmted ia young

aduîts. 1 expected that dda aduits would show kger costs unda DA conditions since

tbyuepreraimdtobovef iwap~rrsouragavai labktothmthandoy~g

~ 3 : M t t p o r i a ~ u d d d . ~ 122

duits, but this did not occur (see Figure 5). Monava, d k t m h g tasic wats did not

M k in Expmbnt 6 depondine on the I d of fk0at.l lobe W o n of puticipants (see

Fisurs 6)-

Fimm 5. Mean percentage deduie in ~cnuicy rate &om singie to divided-attention (DA) fôr each condition in Expakmt (Expt.) 4 and 5, in young and oider adufts.

i Hiah hntai Low frontal

animacy odddig it

DTsfracەng Task

-6. Mernpacentegedecüm~acnincy~tenOmsingletodMded- attention in older a d t h with nbtivdy high versus fow hntai lobe fhction for eachdismctiagtask.

I F t h s s 6 c e i b b ~ ~ ~ ~ o f ~ p m c a P P a i g f e ~ 0 ~ ~ c e s , t h e P

those with lowa lm& of hntal M o n s b d d have showed greater costs. As already

mentioned, my chosen d&acthg tas& moy not be seaative mou@ to differentiate

dinemices m ~vaüabte tesources in ywng and old, and those with felativdy poor &ontal

lobenmction Inotherstudieswbaeuiagediffaaicemdiritrsetingtpdrcostswas

observed, the depadent m u a i n wu latency. H.d I eniphasked the importance of

rnaking l z s p o ~ quickly Q emphsized to prrticipmts the imporrrrît of ouwacy rather

than RT to fespond), age differences in Experiments 4 and 5, ad group ciifferences in

Expaiment 6 might bave emerged. One shouid keep in minci, however, that the costs to

the disLtacting task were always sigiiincant, thereby illutrathg tbat retrieval does make

resource demamis, though the present Qcpaiments do not dow us to ascni these to the

fiontai lobe.

It couid be that our distrachg tada wen much easier than those useâ in other

stuciies (Anderson et ai., 1998; C d & Mdhwd, 1987; Macbt & Buschke, 1983;

WhitingBtSmith, 1997),wheretherewaeagediffkencesindirdnctingtaskcosts

inamed by the DA condition, But gour tasks w m essis, then the &ea ofthe animacy

tad i on mernogr shouid also have been smakr. hsteab its effect on memo ry was quite

substarrtial*

Thescnsuhs,togethawithotbasthushowedm,&eetof~onniem~ry

uada DA at mieval (Andetson a al., 1998; C d c & McDowd, 1987; Macht & Buschke,

1983; Nybeig et al, 1997; Park et al., 1989; Whiting & S e 1997), suggest tbPt

. * * dtmtrnshcdaVBilZLbiiayofattesijidnswrasisaottbeGritidficiorleidiagto

dianrptions in memory recovery under DA. Also, as suggested in p m i o w work (Nyberg

~ 3 t D A & y l o o y u d d b u k l h 126

et d., 1997), the tindiag of iatafaeiux effects simüar m magnitude betwan young and

o k aduhs provides a0 support fm tht Dotion thrt agerelated defkits in episodic

manory am nlrted to reduced cittentiod crprotYcrproty niit the pattern rad magnitude of

i s t e r f m on memory, fiom atba DA condition, did not difféi in Experhnt 6

dcpendine on fml of frontal lobe h c h n provides fiirtba support for this claim.

at cm these resulb t d us about mernom in oldw aduib?

Consistent with the composent-pmcess d e i , whahr one has deterioration in

FL or MIL h a i o n does not ampiify mcqti'büity to diwiption by DA; and wMe poor

MTL W o n d a s lead to pooier meniory, it does mt lead to greater susceptiiility to

dislraction, The data fiom E x p e h a t 4 and 5 suggest that wbat declines with age is

encoàing d o r storage of new information (Morris, Gidg & CrpüS 1988). The ody

differences in pafonnance between young and older aduits wu that olda adula r d e d

fewer words when givm the singie enadhg phase in m e n t 5. The pattern of DA

intaference on memory, however, mrmhed 9müu to young a d t h in ExperUnent 4.

Beuwe the différent encoding conditions in Eqmhent 4 and 5 yieided diffwent obsoiute

levds o f d in the two samptes of ofder d i s , these data are more in iine with the idea

tht aghg is assochted with poorer eacodiag W o r stomge (CRÜC & Byrd, 1982;

RabiiwitL et ai., 1982).

In tamp of the componmt-process modd ofnrrieval, these data togetha with

otbers which fouad no age-reiated krease in the effeds ofDA at reüîevai h m no*

verbai distnctmg terlrq suggest tht retnd on rssoattive tests of memory, ppoc#ds

Oba&atorityudisnLtivdymYmwtodisniptio~~. rtisodyWhenthemem~eytasksbPres

thaspmerrpnsemstio~systemu,the~tuLtbrtIirgeeffedsonmemoryare

ofthc mmmry trace, m the postaior neocortex, Rtba thm in the regions mediating

c 0 n t r 0 1 ~ ~ l l ~ t i k f o r c o o ~ d u a f r t r s L q m t h ~ a s l l l o k a Beause

the ~~ regions are relativdy presavcd with agbg, rad access to these representations

occuril reiativeiy autornatidy 0.e. it rsquueS féw c o p i h fesources), olda adtûts show

a prttem ofinterference effeds simüir to Young duhk

IwillbMflymiewthefindiagsofthea<pauiientrin~thesis.NortIwül

diacuso other worL tht supports the notion tht repmentations in the brain, for difftnat

typesofmrtalll, uecüshct IthenconsidcrinddtheresourœaccountofDA~ea~

atretried,and~evideactigundthisview. Themiplicationsofthisworkwith

respect to our aimnr undastradiog of memory dadllia aswckted with aghg foiiows.

F i , I &mss the present thsis in relation to cunent wok in nairo-imsgiiig. and

propose futun studies aimed at specifyiag the component processes involved in retrieval

of e p i d c memones.

Sammarv of Ex~erimen@

DernonSttation within a liboratory setting of a debilitating &kt of DA on

r e t r i d bas ban VariOble, and sometimeS difficuit to ~chieve. These atpaimaits wae

designed to investigate the d c i p a t e d smp4 aad variable, effects of attentional

maLupuMons on retrïeval. Recmt work by Fernandes ad Moscovitch (2ûW) showed

tht a verbal CUlDIUlng recognition task and a word-monitoring distractiag ta& perfonned

collcucrentiy with led to an approxhbe 3û% dechne in memory pediormnce,

h m fiJT to DA condlaôns. These resuffs s t a d h confrasf io other studies o f v d

ummry in which r e t r i d was p a f o d concurrmtly with non-verbaI dhradng tPslrg

such as card-sorthg (Bpddeley et ai., 19842, digit-monitohg (&mandes ad Moscovitc4

2000), or a visuo-spstiil &sic (Cr& et ai., 19%; Naveh--amin et aL, 1998), which led

to smPn i n t e r f i effects on memory. Based on these data, Fernendes and Moscovitch

(2000) p m p ~ d thit ofmem~ne~ is aot dicnrpted by -rial mqmaiilrh'on~,

k i t t h t ~ h m D A r t ~ u i s e a ~ m c o m p e t i t i 0 u b y t h e d i s t r a c t i n g u e k

t k o o i ~ w o n ~ ~ ~ i l ~ e ~ ~ f e c ~ v e r y o f t h m c m a r y

trace*

I i i i i t i i l ly i iwest ig i tedwhdha~a~~tohoveamemorycompoaeat in

the dbtmdq tuL to produce intedhnœ* Au of the vu=ôal distractuig truks fkom

Fernandes and Moscovitch (2000) had a memory t o i 4 which left open the posstMity that

iiaedkencc effecrcl couid potentidy k âucto conqdtion for maaory structures ntha

thn fm a coaimon, word-based repmmtatbd system as thy suggested. Despite

elimlliatiag the memocy componeiit of the distracthg 1 found that on-üw word-basecl

tasks such as Mimyr or syuable judgements, paformed conanianty with retnevai,

produced large interference &ects. Furthmore, these effkcts wwe not modulated by the

level of processing requind in the distracting lssk. whether semantic or phonologid.

This suggested to me that it was the word-fonn or phodogical aspect of words in the

dislraictuig task, nther than thar semantic content, tht accounted for the large

in- e&ct on msmory.

Consistent with this suggestion, Expaiments 2 ad 3 showed that evm a

&mcting tult invoivhg noasaurawordq thet bas m d c content, produceâ

comparaôb brge enicts on memory. A -c &trachg tult invohriag pi- on. the

other hanci, with no word-fom component, pmduced a signifiant, but d e r , &éct on

memory. hxeasing the 1 4 of di.ffiaiity of the pi- ta&, h m E q a b m t 2 to

E x p a i w a t 3 , d i d w t ~ t h e s i z t o f t h e ~ ~ e c t , Thus,tohawtbetargest

effecton~tbeseacperimeatssuggestthrtthedUeP*ingtPsks&uidengage

word-form or phodogîcal pco~esses.

Anotha miportant f ~ g h m Chapter 2 is thst uxxincy rate on 1 of the

~ ~ ~ ~ U n d ê r D A ~ i t ~ *bmnith€nil

et al. (lm Anderson et al. (1998). Naveb-Benjamin et ai. (1998), und Fernandes and

Mo~covitch ( 2 0 ) , retneval Y not an automtic p~occsq as it draws nswrces away fiom

the dirntçtine ta& unâer DA conditions.

In Cbipta 3.1 d d d an aitemate rccouat ofthe large memory costs nom

Smittmitarialinadistmhgtask Iinvesb'ptedwbetherareductiorhgeoenl

attentional tes~wces, such as those d e d h r frontai Ioôe Wons, wae respo~eible for

the Iarge & i of DA at retrieval. The resuits do not support this interpretation, In

Expaimsnts 4 and 5, young and old aduits were smiüruly affecteci by the DA conditions:

enimPcy dcQsions to words produced much Mer digniptions in memory thsn odd-digit

identification as the distracting Esslc Older aduits did wt show interference e f f i on

mcmoty that were ampMed relative CO tbore obKNed in y o ~ g ad* which would be

expectedbasedonanxwceaccoullt,

As in previous studies (Anderson et ai., 1998; Mach- & Buske, 1983; Nyberg et

ai., 1997; Park et aL, 1989; Whiting & Smith, 1997) DA at r h e v d with a wwverbal

dirffncting~didaotpmducevayirrgee&dsonmanory, andthenwasno

interaction wah age. Moreover, ifcornpethion for rtswces accounts for the large effëcts

of DA b m word-bpred distrcctmg t ah , memory mtaference in participants in

~ 6 w i t h t e l a t i v e l y p o o r ~ ~ 0 1 t s b w k S h n n b e a n a m p M e d r d r t i v e t o t h o s e

with better FL fimaion, because they are rltePdy compromised m terms of âvailrbility of

C e s o ~ .

Withrrspecttocoststotbe~tasks,OhenailiJ~mChspa3are

coasisrentwiththewachisi~nfiomCtirptaZtbrtretnevalisaot~rnlutomsticproceaq

dhd--bk---ontBe

dbmcthg tasks when paformed c o ~ . Howewer, whether these cos& are

sedive to iwailability of generaî pmœshg re- &ed for fkontrl lobe hction, is

questionabk. IfIf were, Wae,thaitho#withpoorerFL M o n diould have showngreater

costs tban those with better function, as tby wouid cequire even greater amounts of

resources to arry out fiontai lobe fiincrions. Thac k scnne indication, b a d on the

resdts ofChpta2 and 3 of t task-spedic componcnt moddathg the size of interference

on disüachg trJc @ormance. That is, similor to the effécts on memory, the use of

simürr materials in the wacumnt tasks produces larger mterfetence on distriibiag tasic

paformance, corn@ to when di,osimilnl. materiais are used.

However, there are âaîa that argue agabut a rnafetial-speafic accowit of

distractMg task costs. F i the ske of the costs to the word-anîmacy and picture-Sue

tasks h m Chapter 2 are similir; brPed on a matd-spcgfic account, one wodd expea

wsts to be greater in the case of word-raimrcy. Wbnt is wm. pmious work by

Fernandes and Moscovitch (2000) showed thai costs did not différ for a word-based

âh t tadq tislr compareci to a digît-based one. Thus mstaia-speaficity of the concurrent

tas& miy contribute to the si& o f k di&a&g fssft cmts, though t does not have as

c o d e n t of an &kt as in the case fm memory interfer;ience.

~ d o t h e s e ~ t d u s a b w t h a w n t r i e v a ( o p a a t e s ? Theysuggesttht

reLneval is mt an obhgatory pfocess. Perfomhg a word-based ta& concurrentiy with

d o f ~ ~ ~ & c a n m t a f e n w i t h ~ n o f t h t a c o c o ~ r e p r e S e s t a t i o m n e c e m a q

hrrecoveryofthettaice. Competitio~iscrcitedmthememoryanddistractmgtub

cocnpcte fbr the same rrpnseaiitionsl systenn,

'Iaece apaiments a h allow one to determine more pnQsdy the Ioau of the DA

a&4 on m u ~ r y , by spoafying wht type of rcpreserdrdion m the distrcaing tulc

ptoaicerr the most htafbaice on the memory task. Bcause the non-word distraaiag

ta& with no d c content, war ahown to disnipt memory, this suggads tht

phomiqical or orthograpbic pn>csaring is mwmry to retneve the words f5om the study

k ~ i t L ~ e t h t s a m a a t i c p n i c a s g a e i n a ~ ~ w i t h o u t a w o r d - f o m

component, a h htedered with ntriewai, it did ao to a letuer d m . Thus, cornpetition

nom a connimat task for semantic repmsmtations W not the p i u w y locus of

. intafaence. possiôiy because semantic repmentations of target items are more

distiihted (Muth, Hutby, Lalonde. Wiggs, Chai et al., 1995) than r e p r d o n s of

word forms*

Recedy compkted work (Moscovitch, Fanrndes & Risalic, in prepadon)

investigated whether word-fotms (ortho&raphy) were noasssry to produce the large

&kt, or whether picture Btimuli couid produce the &èct, when phonoIogid procearing

wrsaisompired. Wefd~interferenceeffeds,onmemoryforawordü~when

participants s h u h m d y made phonoiogical decisim (mie judgernents) to

Soodgnsir rad Vanderwort (1980) Ose drawings. The ske of this e&d was much hger

thn tbit obsaved h m pi- dhtmchg ta& in Experients 2 and 3, in which

participants mada mhacyor size judgements about the same pictures. This sigeests that

the locw of DA &kas a retrievai U at the level of p h o d o ~ representatiom

~thesrmtiines,wefouadtbatwhmtheprocegPingrsquuedfortbe

~ t r o L d i d n o t r s q u i r r p b o l i d ~ ~ p r o c t s P p g b u t ~ a w o r d - f o r m , the

mtaaaeace~oliniemorywrrmmmiil Thatywecompuedtbeintafaeace~

t h # & w t 8 i o ~ ~ - t ~ - & w # e ~ w ~ t o

pmsmted in numerid fonn NQtha ofthse tasks producd iarge dimptions on a net

d test (the deciine in l l ~ e m ~ r y unda both DA maditions was about 10% compareci to

nin attention conditions).

An alternative account of large enrcta h m DA at ntnevsti is thit pnor to output,

words h m the study list are heId in a short-tem memory Wèr, or working manory

store. PnNessmg of i n d g w d s for the dktmchg ta& w y dso require working

memory, Ieading to cornpetition. Ewe consida this working memory MW as it was

desaibcd by Baddeley and Hitch (1974). it is ody verbal-based distrscting usks that

should interfère with rem& whereas sprtiai-based &trac@ tada can be M i e d out

using the visuo-spatial sketchpad, thaeby eiimùinting any cornpetition for workhg

rnemory resoufces. According to this uwwmf one would expect interference to occur

oniy when the disrrscting task requires phonologhi p&g, as this is the code useci by

tbis ahve system to process verbal materid

Howevex, there are two pieces of evidence against this inîerpreîaîion If

mtezference anue because the worlMg memory store for vetbal materiai is overextended

d e r DA cornons wlth a w o r d d a s e d c o ~ task, orda addts shourd have shown

arnpBed Merence Seds in thb c o ~ o n . According to sahhuse's (1996) Wew

@ng b asmhted with 8 cognitive slowhg which redtlces the amount and q@y of

information s h m b m d y avaîWe in working memory, or the " d y d c cspacity. @.

4ûQ ofwo~mearoryry Similarty, Hhskrad c o ~ ~ & zlcb, 1991;

E b u n & Huha, 1992; Hasba, Stokzhis, ZlCEr, & Rypmq 1991) suggest dder addts

hve an mipùmieat in inhEb'ition, lerdmg to a more "chmaed* worLiDg mmory. Givea

effects, manory for words flom the shidy üst should have fecovered. In other work

(Fernandes & Moscovitch, in praparation), 1 tested whssha words fkom the study Est

were üuiy unavaüable, or simply inlrmPlhk. 1 gsve participants a recognition test

foUowing each DA condition, that hcluded ali ofthe words &oom the study list.

Recognition off- m o n enviromnaitil support than frrt d ( C e 1983; 1986; Cm1

et al., 1996). By providing participants with a powaful externai aie (the study word

itseif), I could detamme whether DA intafacd with the trace itself(unaV8ilaôle account),

or with the gemration of appropriate cues to aid i d v q ofthe tiace (iuiaccessible

account). 1 found that there w m more additional words recovered usiag thk test of

mawcy, but that memory did not -ver to the levd of @ormance bbwkg fiill

attention Alsq participants made mon tiilse rLrms fbilowing the anhacy DA condition

compmd to the &cidigit DA coaditioa B a d on there nsultr, 1 believe that the

memories are largdy unavailable folowing disruption b m DA at r e t r i d 1 do

actnawledge, ho-, thet an output mUrfaeaa r#xwn may qpty tpartiupants do

wtatteqttocmywtthernimryandnc9taskriimuaoaeousty, rindmsteadopt fora

PsL-switcbIag stmtegy, tbst waJd dow them to -ver h m tbe DA Hed,

~ E ~ e s r y ~ V 8 i b O C ~ ~ ~ ~

another group f d similar m 8 f d - s p S c &ects of DA at r e t r i d . Robb'hs et ai.

(1996) f d that memory r a d for the arrangement of chas pieces wu, a f k t e d much

mon by a vimo-spatial concurrent tasic thn a verbaldculatory trdc They did not,

howcver, aamk whctha thae wu a recovcry of memory foUowing the DA conditions,

nor did they examine whaha equident e&ds wae found in a population of older

adutts. As mch, this study does not dow one to detainiae w&tba interference was due

to cornpetition for npresentatiod systems, in the smse desaikd by the component-

procesr modd, or for a cornmon 'dave' system in Badddey d Hitchls mode1 of working

-rY-

The Iack of a large intafere~l~e &kt on memory h m digkbased distracthg

tssks (Fernandes ind Moscovitch, 2000; Expriment 4,5$ ofthis thesis) suggests that

digits are represented independently fiom words or word-fom. Consistent with this

ciah, Allison and colieagues (1994), ushg dectrophySologid recordings, showed a

negatbe potenu N20, in discrete regions of the Worm and iderior temporal gyri that

were in Merent locations for face, Ietter-string and numbef stimuli. This led the authors

to conciude th îhere e&ts d E e r e ~ ~ f "moduces" for the procesPiag of numbas, in addition

to tbe prenious suggestion that t h e are sepsrate pn>cessjag streams for faces anci words

@ad, 1990, Farah, Rlfson, Drain, & Tliulq 1998). Interestingiy, they fouad no

o v e in the Iacstion of fsce and letter-string "modules", though in =me cases I e -

saine anci nimkr N2ûûs wae fecofded from the s m e I d o n , d c h mggesteci to the

authors that these moduies may k les3 ibdon9y and sp.otrlEy discrete.

The notion of separate modules go bandie p r o d g of cliffaait types of materials

~ * ~ e m p o s c a b - ~ ~ Q ! = & f ~ > . ~~ thatthebrahuadergoesaproipzsavemodukntyovatmKridwithinaascd~ence

with diffîasat types of mitaials (sa a h Fan& 1998). This s y s t d c npmtmtdon of

knowtedge aîiows one to proccss darses of stimuü, suc& as fiaas and letters, more

efficiently by having certain d e d i d regions of q x d ï d o n I fwe consider the resuits

of thb thesis in this iight, a conse~uence of such moduLnty h hcreaseâ interference when

themoddeisrequirrdsbmbmw&fwtwo~. AnhtaeStmgbofinpuiry

stanming fiom thk work w d d be to coaada whctha young chüdren, who perhsps have

not yet developed distinct letter and number modules, show iarge interference efFécts

underDA~retrievalframbothdigitandword-baseddistractingtasks, ratherthanjust

h m word-based tasks as in addts, who have developed sepamte modules to represent

theae rna tds .

The idea of mal* repmentations D raniDiscait ofKuubourae & Hicks' (1978)

suggestion tht humpari have rnulti- ptocessors and that it is ody when the seme

ptoces~ors are inv01ved simultaaeousîy in two tasks tbst pedormance wül be degrsded

(see aiso Moscovitch & Klein, 1980; Klein, Moscovitcb, & Vigna, 1976). Specificall~~

KinsbounieandHïcicspropodtbatt6e dëgreeofinîerfëëceEom~dtaneoustrsksis

an Unmae fiinaion ofthe "fbctional distance" berwe4n cerebral control centers. Appiied

to the prcsent work, word, nmiba ud p i ~ a m m a t d miy nqUe refativeiy distinct

coatrol~tbuscankproasoedindepededy,~toîitt ieinterfkenceunda

duaI-task conditio~~~,

Otba work has suggested that memory r e t r i d aumot be pediormed d e r duai-

tosir OOIICiitiOnsdueto respoase-selectiundemdsofa seconQry(c&ra&q) EssL

cxwigittgic Taa-despitt*ww*-a'Flrdr*the)i

were eqwPl mmrcedcninnding. It may k that word-based d h w t i n g tasic3 Md

more uadaDA conditions with rdtieval kcause participuUr engageci in more monitoring

of recaii output, and postacphoric piocesaiigpiocesaiig

The variation in Jize of distnctiag tasic mats, with type of mataial in Expairnent

4, cwM be istapreted as evidencc that word-based âktmdq trrks have their effect on

memgr because of cornpethion for attdoual resources. This intaprrtation is m line

with Craik's (1983) proposai that retrievd procesaes arc s i m k to those involveci in

perception. For arempk, th- is an matrre in RT to seamh and h d a target viswl

stimulus when it h plsccd uwng distractors that dure attriiutes with the target (Le.

color, motion) @esimone & Duaaa. 1995). The tncreased RT is taken to mean thaî

w>rs attentional tesources an need to @am the ta& in this case reiative to when the

target b mong dissimk disttadors, wbae the target appeus to "pop out" &om the

distractors without the need for attention Ifcüsacctmg tasic costs dect attentionai

nrourceq the iacrt.sed costs on the uYmrcy ta& rahm to the odd-digit dissacting

ta& that the sarne phenornenon applies to memory: greater attentionai resoufces

are aeedad to retriewe the target wordIist w6en puticipurm are sllnuTtaneously pcocessulg

siniilu compcued to dirmmilu matezhîs.

Tho@ 1 ackncswiedge tht iaaeued am@&m for &ternional resoutces mry

wntrikitesomew~~,ths~eiataâxaCeeatictsoammorypaformaace,it~ot

be the oaly source of coqmition Eit were, olda ac&its m Expallnan 4, and the olda

aduhshExpkmt6, c l a s s i f i e d a s ~ ~ p o o r e r a w C a l l o b e ~ o ~ wodd

b.vb---~it-ZLIIQ-trrok@-uada

dual-taskcoaditioaswithramircy~task.

In tryine to accouat for the b e intedikence &eds ofDA a retneval, f?om

w o r d ~ mrtairl, it U worth considerhg otber wwk in which memory &ml W

dLNpted In a study by Brown (1%8), participants studicd 25 ofthe 50 U S mes,

folowed by a r e d attempt of iU 50 states. Rdrtive to 8 control group that did aot have

an initial sbidy &og thy d d more &the d e d 25 strtes; but, wbat is interesthg

is tbat they d e d fewa of the unstudied 25. 1t a p p e ~ that study ofpart of the üst of

mes had Uihiided recall fiom the complementary subset. This pheaomenon is réferreci to

in the iïtaohtre as the put-üst cueing e&t, whae the act of retrieval inhibits mai l of

iafonaation that is assdated with the succedûiiy retrieved tuget information (Rmdieer,

1973; Slamech 1968).

In more recent wok Anderson, Bjork and Bjork (1994) asked participants to

pmctice retrieving hlf ofthe items fkom each of severai categories. In an ensuing

d e m i session, in which aü items h m the categories were to be d e d , they fouad

tht recail of non-practiced items was inhirted nlrtive to a control condition. The

uithon explein this in tams or hEi16Itt'on or suppressiin ofu~lp~iccd items d h g

the rarievai pirctice session that pemkts to the subsequent M e v a l session. Tâus,

rezrievil in pert-list cueing arpaimepu acts to fhState the recell of mget item by

suppmshg issoa.ted, but 1~111-target h. An kasthg fiitun study w d d be to

investi&atewbcthar~nmerintafp.eacea&aswwldkfOuadwbeatheshidyüst

coatrinedrdrted.nnnurUIUreIBfed~Etheccisiwamwn~gprocess

~ t b c ~ m e m o r ) t a 4 n c t r ~ k ~ ~ & D & . a d t b e ~ - l i a t

cueiq &kt, reûievaf may mbi ibd more fw the bt of hi& uuonrted words.

Wbile the part-üst cueing e t h t has bem obsaved fbr memory of weii-kamd,

d c idormation, ad categorized word ib, the meanir by which retrievai fàüs may

be tdated to the interference &ectS obscwed in this thesis, for a Iist of unrelateci words.

For example, in rlmoart di of rny slrpssimaits, the MQCL~C~ effbct on memory passted,

when tbe dud-tasl coadition in gsaaJ prctiapants were mabk to r d

additionai words h m the study iist, at the completion of t&e DA sessior It rnay be tbpt

unda DA conditio~w with word-based distracthg tasks, d of target words wur

possible oaly by suppteapng the sniilrr, dkacthg îask items. Whatever the n d

mcchoasxn for this suppresüon, it may have oaended to whicbmr target words were not

recalled duhg the DA wadition. Tbis wodd accwat for the persistent éffects of DA on

memory. Tbis aiso suggests that whai retrieval ocnus under conditions in which one is

fàced with havhg to simultaneously ptocess Oimüpr materials, some amount of suppression

mi@ be mxswy in orda to d any items.

Insofbr as the process9ig ofomiilaf materiah is concemecl, theories aimed at

unci- seidon of target items mmg diatnctorsare Bnportimt to cornida.

Rathcr than s u b m i i to a suppnssion vxount of sdection, Lakge (1997) suggests that

attention to target Mfodon is enbsnad, W topdown ineuences firom tht PFC. These

M o n to enhance corticai col^ in the posteaïor p M d cortex (PPC), wbich

repttsena the "object" of attention, by the action of the thahus, whïch speQnes the

immsity of BCtiVatiOnofneuronsintbePPC. Intherbeenceofsupprtingsignrlsaorn

thePFC,adivatiminPPCdccays. Intbaareofthirthesis,theuobject"maykwords

eech**iisl ~ t 8 o t c i A a n r ~ ~ a t r i c u ; r l , o a c c i t A n b i a i . L ~ t h

PFC miy act to aihuice the npdons of words, stond in the PX. Under DA

conditiom with word-ôased distractors, the d e d o n of C O ~ U M ~ ~ Q to enhruice may be

mipUiad, o r t h c l d of- reduced, asrttsmionmust aiso bedirected to the

â h d n g task. Retrieval unda DA ushg dissimüar materid is not subject to these

mphats as items fkom the target word üst may be abject to a preaîtentive "pop out"

cnéct (%ismm & Gdade, 1980). Thedrawback of thir iccouat however, U uut it

phces the locus of the Metence &kct witb a homuncuius in the PFC, whose role is in

pucehg out attentionai fegoufces. Also, as aircady discuared in Chapter 3, a d in the

p d o w section felating memory to perception, the !ack of ampiified efkcts in older

adults and those with poor fiontai fiuiction are inconsistent with this account.

A Mer line of iaquiry stemm@ h m the component process mode1 imroives

showing that the material-@c e&ctr of DA a! rCtnevPl appiy to other tests of

memory. Zn thais 1 considered ody h e d tests of mesnory. Accordhg to the

mode& ifencds Mse h m coqethion for -tuai npresentation reaponsible for the

content of the memory, then the material spedic effkcts reporteci hae shodd &O be

observed on recognition, Pnd even on perceptuaî hpE& tests ofmemory.

Wth respect to recognition tests. 1 have conducted some preliminory

irwesb'gations, and the resuits ue inhwith the oompoocat-process modd ( F d e &

Mogcovitch, m prepadon b). Recognition trds O& more umviroamntal support" thaa

dtestsofmem~ry: rnodanalcue(theworditself)cinbeusedtoaidreüiewf, in

comnst to &ee rrall. in which iataiiP1 aie g d o n h d e d . (Craik et 11,19%;

A n d m et al., 1998). Along the same lincs, reco@on may be dmnn more by

a O

~ ~ d l e 4 % 5 0 ~ e O B f r O C f 8 0 8 6 e i O C t e ~ p f o e e s s a s ~ , &!BQ. E

word-bascd dbacthg wks interke with iataail nibgeaeration, them we should see

1- interf~te~;e whai a recognition ÇIJr h useci, Sooe recognition o E i extemai ares to

guide r e h n d . Also, tecognition d e s more heady on Maior parietai activations

compued to recaü, which telies more on casMIar-frontal pathway activation, beIieved to

refled the W o n of cuagenemtion needed fw recaii (Cabeza et al., 1997). IfDA

at retrievd interfefed with ccddbfiontaf activation, thn performance on î t rccogmtion

test, which does aot d y as heavily on this circuit, would nat be affecteci under DA with

word-baseci distracthg tasks. The component-pr- d e i piaces the locus of

htafasace at the Id of perceptual representations, and predicts that recognition wouid

be irnpPIl#i. as is free recaii, though to a lesser extent since recognition is an easier test of

v. In line with the compouent-process model, I found recognition pedionnance,

m d as hit Rte - mse Jum rate was .82 unda fuil attention, -75 under DA

conditions with an odd-digit tapis lad .66 under DA conditions with an animscy task

These d t s show thit interference &kts fiom word-buseci disttacting tasks are mt due

to mippirrd genemtion, or süaîegiCc searcFi 6 r responses MÇe the &a i9 maintained

aven with a rcco@nition test of memory. 1 ah coasidered whether the poorer

patowtaceundaDAcorrditio~wahthcrniavcytdrwrsduetorninueaQeA

~~~~e~frbüitytof&edams,butthUwurnotthease. Theharatewu,0.84utrderfÙü

attention, 0.78 and 0.69 intheDACOûdifionanthtbtodd-di~d ammicytasks

n q e d d y . ~Anoivaiswed~thehitrateradhlsePhnnrrtedidnotdi&r~o~~

L - drtre- ms&qwàd- PO OC CE-^^^ . . . .

crnnot acamt for thematerial-Speanc htdmceefnd.

Ifintafonace ocans for paceptual represcntations, u suggested by the model,

oncwauldsxpecttoscsmatCrint-~mtafénaccDebcts~onpacepturl~licit

tests of memory. Like recognition, paoepaul implicit tests do wt repuire gewntion of

i n t d aies to guide r h d . DA at d e v a l ushg a word-bssed distracthg ta& may

dîsrupt the m o n ofworQ iu manory, a d produproduce interfkmce evca on auch

impliQttestssincethesereiyhavüyoapemeptdsystenis.

In addition to c o n s i d e other tests of mcmory, the compommt-proccss mode1

can be fiutha e v a i ~ e d Usng neuroimaging* Sevenl studies have shown .ctivatons

miluig d e v a l in ri@ prefkontal @A10) .nd oubcofticai amas inc1uding pu- giobus

pailidiig, a d thaiamq as well as in the posterior c h p k , aiaais, Md cerebdum

(Csban & Nybag, 1997; Tulving et ai., 1994. fidaka, Anderson, Kapur, Cabem, &

Craik, 2000). hterestingîy, areas of deaeaPed achtion wae aiso wted in a

neuraimaging shidy ofDA & i on rttriad, in bilateral temporal regions extendhg to

the mWi. cortex @daka et al., 2000). This ôiiernporal negativky, in DA c o m p d to full

ittcntion wdi50as, wiu a b seen m neurohugkg stuâCcs ofvtouat mention (Corbetca a

PL, 1991), Md short-tam memory (Andreason et al., 1995). This negitivity b been

interpreted as ta&-nlsted inhiiiition h m otha bnin regiong, to prevent proœssing of

idevant mamiai (Grasby a IL, 1993; Nybas et aL, 1996; Raichle a ai., 1994).

This pmposition briagS up the parsibiüty of an aherriative locus for the large DA

enectSobsavcdhtbisthesis-,thaMTL~y~~ntributetothtmtafénnce&ect~mDA

uSpeworb6asedd&ra&ng~. ItispossobIethattbenetworkpmemofbrainrctMty

ib-assadaDA-**v----

maeL#d siniiluity ofdd-Esdrs. patups there is more inhiitioa in MTUH regions, a by-

prom>ct of- to prrvesit pmcesshg of w)*relevant m a t a This possibility wouid be

d to Mvestigate m the future. The component-procesis modd aiggests that

mtafaenacanoccurat t h e I e v e i o f t h e ~ o r ~ p o s t e r i o r ~ r t e x , ûnthebasis

of tbe arpaimeau in this thesis, ud those m FefllZIides rad Moscovitch (2ûûû), 1 fivor

theneocortexinterpretPtioanthathana~loatoforthe&ea ThusIexpectthat

interference Mses fiom cornpetition in posterior @oas tesp0nsiiIe for the content of

memory traces, iikeiy in parietai d o r pnainau, regions.

Rdèrences

A k t , M. S. (1988). Gamû haaies in gaitric nsuropsycbology. In MS. Ama &

M.B. MOM (Eh.) Ganitnc Nano~sycbot~gy @p. 1-10). New Yorlr: W o r d .

Allison, T., McCarthy, G., Nobre, A, Puce, A, &Be@, A (1994). Human

e x m s t r h viswl cortex aucl the! -on o f F m Wo* Numbas, and Colora

5,544554.

Aliport, A (1993). Attention and contml: Have we beai asking the mong

questions? A criticai miew of twenty-fie years* In DE. Meyer & S. Komblum. (Eds.)

on and Pafonnance 14: Svnermes ui acDauamSal p ~ n . . bol-. a r h f i d .

intelli~aice,

d conntb . . a

ve neurosmeflce. Cambridge, Mk- MIT Press

AUport, D.A., Antonis, B., & Reynolds, P. (1972). On the chision of aîîention: A

225-23 5.

Anderson, MC., Bjorlg RA, & Bjork, EL. (1994). Remembering CM cause

forgetthg: Meval dynamics in long-tam memory. Journai of Expaimental Psychology:

&8m& ItkmFy, re8 c- 2@, 106-3-108.

Anderson, ND., C e FJM, & Nweh-Benjamin, M. (1998). The attentionai

deman& of eacodmg d retrievai in younger rnd oIder ad* I Evidence h m divideci

attention costs. Psvchoiogy and A a u r ~ 13% 405423.

Admason, OZeaty, Amdt, S., C W o , T, Eiîrîig, R, Rezai, K, W d c h , GL.,

Boles Ponto, LL7 & Bicha- RD. (1995). Short-term a d long-tam vabsl memory: A

ad- 146

L

positron mission tomography shidy. Proceedinns of the Nationai Acadagy of Science. 92.

5111-5115.

Baddeiey. AD. (1986). Wokhg Uemow Oxford: Oxford University Press.

Baddeiey, AD. (1992). b workïng memory working? The Meenth BartIett kdure.

Journri of Expehental Psychology. 44A f l), 1-3 1.

Bdddy, AD.. &ES&&, G. (1974). Working meraory. In OH. Bower@d.), Tbp

psvcholopv of leaniin~ and motivation Vol. 8 (pp. 47-90). New York: Academk Press.

Baddeley, AD., & Wilson, B. (1988). Frontal mnesia and the dysexecutive

syndrome. B- 212-23û.

Baddeley, AD., Lewis, V., Eldridge, M., & Thomson, N. (1984). Attention and

retrievai fkom long-tan memory. Journal of Exbecimentai P-hology: 1 13. (4).

5 18-540.

Bimbpun, LM., Parker, E.S., Hartiey, J.T., & Noble, ES. (1978). Alcohol and

f Verbal Leanruin and memoty: Retrieval proasses. J o d O Vabsl Bdiaviour. 17.325-

335.

Bro&&.R (1963). T h e ~ o f ~ b y r r o d i a g Ouarter&Jaumal

pf Povcholopy. 19,289-299.

Brooks, L X (1968). Spotlll and verbai components of the act of M. c.00di.a

Ja>nirl of Psvcholow. 22 (51 349-368.

Brown, J. (1968). R e c i p r d fiiciIitation and impairment of& recaîl. Psychonomic

Se- 10,4142.

C a k a , R, & Nybers. L. (1997). ha&g cognitio~ An enipiricaI rcview of PET

stuclies with n o r d subjects. J o d of Connitive Neuroscience, 9, 1-26.

~R,Gndy ,CL. ,Nybag ,L . ,McIntosh ,A .R . ,T~E.K;rpur ,S . ,

J~M@s, JM., Houle, S., & Craik, F.1.M. (1997). Ago-related difkrmcr in n d acthity

during memory enadhg and retrievai: A positron enhion tontography sbdy. Jwmal of

Neuroscience 17,391400.

Carrier, LM., & Plshler, H (1995). Attentional Iimits in mexnory retrieval. Journai

9f p n P . . 21, 1339-1348.

CoEey, CE., WiIke~ls01-1, W.E., Psrashos, LA, Sordy, S.AR, S u b a n , RJ.,

Fatterson, L.I., Figiei, G.S., Webb, WC., S p b , CE., & Djm& W.T. (1992).

Quoatititive cerebrai amtorny ofthe a& bnin: A cross-sectionaf study using magnetic

monance ha&. 527-5536.

Corktt., M, Miein, FM, Dobmeyer, S., S&ilmaa, GL., & Petersen, S.E.. (1991).

Sefsctive and fideci atîention duriDg visuai dwcrmunati . 0 0 'on of shape, cdour, and speed:

f Neuroscience Functiod anatomy by positron emission tomography. Journal O 11,2383-

2402.

C e FJM. (1982). Selective changes in enoodiiig as a fundon of reduced

pmcesshg crpocitycrpocity In F. 1. HofEma, & 6, van der Mea @ILS.), Cognaive research in

psychoIogy @p. 152-161). B&: D W .

C e F N (1983). On the transEa of infocmation fiom temponry to pemment

hicai Tdom of the Rovd Society ofLodon Sen metn~r~. P~JOSO~J 'es B30& 341-359,

Cr&, F.LU (1986). A W o n a l accouat of age ciiffierences in rnemory. in F. Küx

@p. 409422). North HolLid: Eisevier Science Publbks, B. V.

C e FIM. (200). The effécts of dMdhg attention an aicodiag ad d e v a t

procerses. In K L. M g a , J. S. Nairrse, L Neath, & A M. Supremant (Eds.). The nature

L s honor of Robat G. Cmwda. Washington, DC: American

Psyychologial kcsocinton.

Craik, FJM. Cm PM). Humaa memory and aging. In J% e Internationai Congres of Psycholoqy. Stockhoim, AuguaS 2000.

C e F.IM, & Lockhart, RS. (1972). Leveis of piocessing: A hmework for

f Vabal Leaniing and Verbai Bdwi memoryre~earch. J o d O 'or. 1 1,671484.

Crsür, F I X , & Byrd, M. (1982). Aghg and cognitive deficits: The d e of

attentional resou~ces. In F.IM. Craik & S. Trehub m.), Aghg and cognitive processes

@p. 191-21 1). New York: Plm~m.

C e F I U , & McDowd, JM (1987). Age di- in d and recognition.

C d , F I M , & JenniasS, J U (1992). RumPn Memry. In F I M CrPilr & T A

SaWouse m.), The imdbok of Sgmg and cognition @p. 51-1 10). Hillsdale, NJ:

Edbam.

Cr& F L U , GoMni, R, Na+- M., & Anderson, ND. (19%). The

dk t s of dMdd attdon on amcihg and teaieval pracesres in human memory. J o d

125, (2), 159-180.

&mm, H.V. (1991). BQlZOdllldPiacs, mawry and mood: A reYiew.

&vc)K,~barmacoIogy. los., 1-8

Delis, DE., Kramer, J., Kapian, E.. & Oba, B.A (1987). The Califomia Vabai

Desirnone, R, & Ihmca~, J. (1995). N d Meachnisms o f s e l e c h visuril

attention. RMew o f N e w 0 ~ 4 ~ 193-222.

Dywm, J. & Jacoby, LL. (1990). Effecfs ofaghg on murce monitoring: DiGerences

in mcapti'bÜity to hlse m. Pwolopy uid A&g& 379-387.

Farab, MI. (1990). Vsual inno* Cambridge, M k MIT Pnss.

Farah, MI., Wilson, KD., Drain, M., and Tanaka, IN. (1998). What is "Speaai"

about &ce perception? Psychoioaicol Review. 105 (31 482498.

Famer, E.W., Bemm, J.V., & Fletcher, YL. (1986). Evidence for a visuo-sp.tiai

Ex&med Psvcholom. 38 (4Ah 675-688.

Fernandes, M A & Moscovitch, U (2000). DMded attention and memory:

Evideacc of subadiaiia intedimnœ e&cts at rebievsl and e n d h g Journal of

&&mentd PsychoIogy: G e n d 129 Q'). 155-176.

Fernandes, M.A & Moscovitch, M. (submitted). Factors n m d a h g the Hbct of

divideci attention at rebievsl: maanonics, d c s , p b o d c s or word-forms? J b k n q

d- . .

Fernandes, M, & MoscoMtcb, M. Cm pnpurtion). Persistent &&s of divided

attention at retrieval: unavailable or hccaibie?

Fernandes, M., & Moscovitch, M. Çm prepadon b). Divided attention durhg

recognition

F0Istei.q M.F., Folstein, SE, & McHugh, P R (1975). Mini-Mental State: A

p d c a i metbod for grading the cognitive state of patients for the ciinician. Joumai of

Psyhiatri~ Raiearch 12, 189-198.

Francis, WN.& Kucera, H (1982). Freauency d @ s of Ennüsh UV. Houghton

Mi88Ui Company: Boston

Frieâman, A, Poison, MC., WO~, CG., & Ga&& S. (1982). Dividing attention

within rad between hexnisphres: Teshg a muhiple rrrrwcg approach to limitedaPoCity

fExdmenfal - infimation procdng- Journal O Psvcholopy: Human Perception and

Pd-. 8 (51,625-650.

FueteF, J M (1997). Tkpdcmacoctcx. M a t o m y , ~ W * u d

~uro~@&w of fiontai [ o h 3rd ed Philidelpbia: Lippincott-Raven,

Güslry, EL., Poker, MR., & Rwthi- B.C. (1995). Double dissociation

betwœa k m a d mëmôry. N&?~svch01& (21, 229-23 5.

Gndy, CL., Mdntosh, AR, Homkq B., Maisog 1. Ah., Ungerieider, L.G.,

Meatir, MJ., Pierini, P., Scbsph, MB., & Horby, J.V. (1995). Age+related reductions in

human fccogdon memory due to miprired enadkg SQence 269.2 1 8-22 1.

Gra9by, P M , Frith, CD., Friston, K, Fracko* R.S.J., & Dolan, RJ. (1993).

Activation of the human hippoampll fbR11Btia durhg audimry-verbal long-term memory.

Neutosci- Ietters. 163, 185-188.

Gregg, V H (1976). Word firesuency, recognition and d. In J. Brown m.), . Chichester, Engiand: Wilcy.

GrifRh, D. (1 976). The attentional dermids of mnemonic controi processes.

Memorv and C o d o n 4, 103-108.

Gur, RC., Gur, RE., Obrist, W.D., SkoInick, BE., & Reivich, M. (1987). Age und

regionai cerebd blood fi ow at rest and during cognitive Archms of General

psychiaüy. 44,61742 1.

EIiliim, V.P., & Hasher, L. (1992). Age and nnilrbüity of iaferences. ~sycblogy

IJd- 5 6 6 4 .

BU, K'110~llftl~, L& Wadq lB, & Taylor* Jg. (1988). Modifieci Wisconsin

C d Sorthg Tm& in eiâaiy nord, deprraced, anci demented patients. Chical

Neumpsy&dpoisr 2,49956.

m, A A, d Me, D M (1999). Agarrirsed diffénnces and similarities h

QICcrsL: Journal of m e n t a i PsychoIw G e n d 128 (4). 4 16449.

Hmk, L., & ZiJs, R T. (1988). Wodehg m~lll~ry. comprehension, and a-: A

nvMw ud a new view. In G. R Bower w), The psychology of 1 d g Md motivation

(pp. 193-W). New York Academic Ress.

W. L., S t d t z h , EX, ZUdq RT., & Rypmq B. (1991). Age and Inhibition.

Jourail of Ps_vcholo~ Lemhz Memo~. and Coention . . 17, 163-169.

Waka, T., Anderson, ND., Kjpn, S.. Cabeza, R, & Cm& F.IM (2000). The

effects of ciivided attention on encoding ad retnevai in episcxiic memory reveded by

f Connrtrve . O Neamscience. positron &on tomography. Jocniill O 12,267-280.

Jacoby. LL. (199 1). A process dissociation firamewodc Separating automatic h m

intentionai uses of memory. J o d of Memory and Lan-e. 303 5 13-54 1.

Johaeton, W.A, Gr- SN., Fi, RP., & Ahth, D.W. (1970). Divided

attention: A vehic1e for monitoring memory processesOceSSeS J o u d of Exnerimena PsvcholoPv.

fi (11, 164-171.

Jobiwton, W.A, Grifntb, D., & Wagst& RR (1972). Speed, ~ccuracy, and ease of

d J o d of Verbal Lamine and VeM Behaviour. 1 1.5 12-520.

Knhnmian.Il_ (1973). AttePtionarde Engk0adCmNI:PréntiCeHall.

Kelley. WM, Mi* F M , McDermott, KB., Buckner, RL., Raichle, ME., Coheq

N.J., Oübger, J M , ALkdpk, E, Conturo, TE.. Snyder, M., & Peterson, S K (1998).

Hemisphenc speQsliption in hmun dorsai fiontal cortex and medial temporai lobe for verbal

d nonvttbal memory encodhg. Namin 20,927-936.

Kdlog, KT., Cockh, T., & Boume, LX., Jr. (1982). Comcious attensioiral

d d of aicodiag rad rdri& h m long tam rnemory. American Journal of

95, 183.198.

KUlrbourae, M., & Hicks, RE, (1978). Fundionai cetebrai spoce: a modd for

ovdow, transfer and intediamcc effsctr in human paformance: a tutotiat mKw. In J.

Rcquia @ci.), Attention and Perfomiance W. @p. 345362). Hülsdrile. New Jasey:

r Edbaum Eribaum~ssociates.

Kirlg RE., (1995). ExPerirnentaI Desim: Procedures for the Behavioral Sciences,

Third Edition. BrooWCole Publishing Company: P d c Grove.

KJeh, D., Moscovitch, M., Vgm, C. (1973). Attentional mechanisnu and

paaptuaî Isymmetnes in tachistoscope recognition of words anci foces. N ~ a ~ o ~ ~ h o l o n i q

14 (1) 55-66. -

Kfingkrg, T. & Roiand, PB. (1997). Merference between two concurrent tasks is

assochted with M o n of ovalapping fields in the cortex. Coolifive Brain Resesrch 6.1-

8.

Khi& KT., Gdnwceclry, W., &&abhi, D. (L99S)- Rdeofhuman*d

cortex in attention cwtrol. In Epilepsy ad the nuictioaal rnatomy of the fiontai lobe.

Advances in N~u~o~ow. 66,2 1-34.

Kolers, P.A (1973). Remembaing o ~ m I I S Mmmy and C O ~ O I L 1, (3). 347-

355.

me, D. (1997). Attention, Awmwss, and the Trbgular Cirait.

CoPlltfion I . 6,149481,

Leenders, KL., Peau& D., A A , &stba, JD., Buckinghq P.,

Healy, M.J.R., Gibbs, J M , Wse. RJ-S., EEatazawa, J., HeroId, S., Beany, RP., Brooks,

D.J., Spmiis, T., Rhodes, C., Frackowiak, RS. J., & Jones, T. (1990). Cerebd b l d flow,

blood volume, and oxygen ut&ati011: N o d values and s&cts of age. Praia 113,2747.

Logie, RH (1986). Visuo-Spatiii proaaces in working memory. Quartedy JO-

pf ExDaimaital Ps~1cholow. 3 8 4 229-247.

Lunq AR (1966). Ebpher cortical bct ions in New YorL: Basic books.

Macht, ML., & Buschke, EX. (1983). Age differences in cognitive effort in recall.

Journai of CierontoIonv. 38,695-700.

Uadden, D. J., Turkington, T.G., Provende, J M , Denny, L.L., Hkwk T.C.,

LaurenceX, and Coleman, RE. (1999). Ad& age diffkrences in the bctional

rmmwtomy of verbal reco@on memory. Hw~in% 7.121.1 15- 13 5.

Muth, A, Wigp, CL., Lalo* F., & C. (1994). Word retrievat to letter

andsemanticcues: A d o u b ( e m I . mnormrl-upine-til.qk.p

Neuf~p~choIo~gk 32. 12, 1487-1494.

M;irtu5 A, Haxby, J.V., Lalonde, F M , Wtggs, CL. et ai. (1995). m e corticai

regions urociued with knowledge ofcolor and knowiedgc of doci+ science. 270 -32

102-105,

McLAhd, J.L., McNaughton, BL., & O%&, RC. (1995). Why thae are

complemcstary l e d g systems in the bippocampus and llcocortex Insights âom the

10% 419457.

Milner, B., Petrides, M., & Smith, UL. (1985). Frontal lobes and the temporal

orgamnton of memory. Humui N e u r o b i o f o ~ 137-142.

Msldch, M., & Appenzeiler, T. (1987). The anatomy of memory. Scientific

256* 80-89Motay, N. (1967). What h crp.city limited? A mmey and 4.

Psvcholo&a. - 27,8492.

MOI%', N. (1969). L~&&Q Md aaiti0n HUIIlondsw~rih, &l@d: pai@h

Books.

Mo* CD., Bransfîord, ID., & Fmnks, JJ. (1977). Lmls of p&g v a s u ~

tr~mdk- * ~ r O = a 3 - ~ o f Verbai-V- 161 5 1

5 19-533.

Adonis, RG., Gck, ML., & Cr&, F-LM. (1988). Processin8 nsources md age

. . d ï f k n a s in workîng memory. Memory and Cornon 16 (4). 362-366.

Mo500vit4 M (1976). On the represenWi0il of laquage in the right hemisphere of

rig&t-banded people. BaÛn Md -e. 3 (1). 47-71.

moduluity and episodic memory. &mmi of Cluucai wd FxberUnental Ne~~~~sychology, .

rl.

(2)* 276290.

Moscovitch, M. (1995b). Modeb of W&OUSH~S~) snd manory. In M. S. Gaaaniga

(Ed.). The cognitive neurosciences @p. 1341-1356). Cambridge, MA, USA: MIT Press.

Moscovitdi, M. & Klein, D. (1980). MBterial-spcQfic perceptual interfice for

l tk t ld~dsdâîes: I l n p k m s 0 . ~ i s o d d o o f c s Q P c i t y ü n i a o t i o n q ~ ~ r a d

Moscovitch, M. & Uniüta, C. (1991). ConsQous and noncodous aspects of

maiiory: A neuropsychologicai framewotk of rnoddes anâ centrai systems. In R Lister &

K WQnesrtaa. @ds.), Peyxdves in CoiBplttve . . N~osc ience~ London: Mord University

Press*

Moscovitch, M, & Wmouu, G. (1992). The neu~~psychology of memory and

a&g. In F. L M. Craik & T. Salthow (Eds.)¶ The bandbook of aghg and cognition @p.

3 15-372). HUsdale, NJ: Edbruim.

Momvitch, M, F d e s , M. A, & Prisetac, S. (ii prepdon). Syilabe decisions

to pictures or words dimpts episodic retrieval when paformed co~~lftently.

Moscovirch, M., Kapur, S., Kôhlcr, S., & H d e , S. (1995). Dintind n d

cordates of v i d long-term memary for spatiai I d o n and object identity: A positron

emission tomography (PET) study in humuns. Proceeding? of the National Academv of

Sciences of USA 92,372 1-3725.

MurQJs R B.&. (19653. E & c t s o f r ~ t i o l o n - t c r m ~ ~

Bntisb Journal o f Psvchoionv. 56,4 13-419.

Nd& L. & Moscovitch, M (1997). Manory comlidasion, retrograde d and

the hippocampPl complac ûarrot ODniion m NdioI04yLz 2 17-227.

Naveh-eenJ*amio, M., C e FLM, G w J., & Dai, H (1998). E f f i of divideci

attention on encoding ami fetrievai pmœsses in brrmsn manory: Furtber support for an

1091-1 104.

Norman, D-A, & Bobrow, D.G. (1975). On Asta-timited and resou~cc-Iimited

pmœsscs. Cosmitive PsvcboIogyJ 44-64.

Normun, D.A. & Bobrow, D.G. (1976). On the aniiysis of paformance opaatllig

. &m%em&a. Psychologid Review, 83, SOS5 10.

Nykr& L., Cab- R, & Tuhimg, E. (1996). PET shidies of d i n g and

retrievai: theHERAmadd. P p J 1313148.

Nyberg, L., Tulving, E., Habib, R, Nilsson, L . G , Kapur, S., Hode, S., Cabas, R

E. L., & Mdntosh, A R (1995). Functionai brain m ~ p j of cetrieval d e and rrcovery of

epieodic informationn ÇpgNtive Neuroscience and Neun,~sycholum. 7# 249-252.

NybqJ,., Nüsson, L . 4 , O l o b n , U., & Bitchan, L. (1997). Effects of division

of attention during encoâing and rrtrievrl on age Werences in cpisodic memory.

E 23, 137-143.

Pantano, P., Baron, LC., LebnmGrdié, P., Duquesnoy, N., Bousser, M. 4.. &

Cowsr,D. (1984) ~caebralhtnnAfla~aodaxygm<nnaimaiianhhrnnancigiae

Stroke, 15.635641.

Pa& D.C., Smith, AD., hdly, WN., & Lahm (1989). Effkcts of age and a

dMded attention tasL pnsmted c h h g cncodiag d ntrievil on memory. Joumal of

-gy: = Mmi- and ~ o 4 m ' o ~ . . 15,1185-1 191,

Ray% C. L., Johnson, M K., Mitchd, K 3, Nolde, S. F., D'EspoPto, M. (2000).

&IRL nmctigatiom of left lad ri* PFC eoQtriI.buti0~ to e p w c remcmbaing.

Iopy. 28 (21 197-206.

Rabhowitz, J.C., Craüc, F.I.M., & Ackemm, BP. (1982). Aprocessin8remurce

account of age diffefetlces in recaii. CMdiDD Journal of Psycholonv. 36,325-344.

Raichle, ME., Fiez, J.A, Videen, T.O., ModRod, AUK, Pado, J.V., Fox, P.T., &

Petenen, S E . (1994). Practice-dated changes in himan brah fllnaional anatomy during

nonmotor l e . Cerebral Corta 4,8026.

Robb'i T.W., Andetson, E., Barker, DR., Bradley, AC., Feameyhough, C.,

Hetlson, R, Hudson, S., & Baddeley, A (1996). Worbg m m ~ r y in c h . Mernory &

Çppnition. 24,83093.

Roediger, HL., III. (19î3). Inbiiition in mail Born cueing with r d targets.

J o d of V a b a l d a 644457.

Rwdiger, Weldoa, US., & Chdis, B i t (1989). Explnlliing diSSOCi8tions

between Ïmplicit end explica tneasures of retentiox a prooessiDg accomt. In EL. Roediger,

III & FLiK Cnüc m), V B and -0- EsSap in honour of Endel

Tdving. @p342), msdale,

Sai- T.A (1985). A the~nr of conmtive aniag. Amsterdam: North Hoiland.

Sphhouse. T.A (1991). Theoretr*cal ~ap~ecbvrr . .

* o n c o n m t m a g & . ~ e , ~

Erbaum.

sa ihuc , T.A (1996). The procas@-speed Wry of aduit age diffiences in

cognition PsyaK>logiical Review. 103,403428.

Sai- TA, Rom JD., Prill, KA. (1984). Division of attention: Age

diffierttlce~ on a v k d y presented memory ta& Mew>- & Codon. 1% 613420.

Scbscta, DL (1987b). Mem~v, ~XIUE& cnd f i o d l& dypfuncii~~~,

Psvchobiolopy. IS,21-36.

Shrtllice, T., & Burgess, P. (1 99 1). Hi&+order cognitive impairments ond &ontal

lobe lesions in man. In KS. Levin, & KM Eisaibag (Eds.), Frontal lobe funaion and

dvsfiinaion @p. 125-138). New York. NY: Oxford Univasity Press.

Shimimura, M., & Squire, L.R. (1987). A ~auopsychological shidy of M

fEx+maI - mmocyanâsourceamnesk J o d o Psvcholow LMC. 13,466473.

Siamecka, N.J. (1968). An Qcomiartion of trace storsga in fiee r d . b u d d

Pvchology. 76,504-513.

Snodgmss, J.G., & Vanderwart, M. (1980). A standardized set of 260 pictures:

Norms for arme agmmez& image agmment, f h i k i t y , and visuai complexity. Journal of

S p m O., & Baiton. AL. (1977'). N-N Center Com~rehensive

ExamiMtion fbr ADW (NCCEA). Victoria: UnkaSay of Vktoria Neuropsychofogy

-w S e LR, Cohen, W., & Nadci, L. (1984). The medlll temporal mgion rad

memory conwlidation: A new brypothesis. in E+ W- & E. Parker (W.), Memorv

nsdlidation @p. 185-210). EIiihdde, NJ: Eribgl~n.

Stuss, D.T., Eskes, GA, and Foster, JX. (1994). Expaimeatai ~uropsychological

studies of fiontai lobe functions. In F. B o k & J. Grafman (Eds.), Handbook of

NeuropsychoIogy @p. 149-1 85) Amsterdam Elsevier.

Tdsnin, A M , & Gelade, G. (1980). A fClture integration thewy of attention.

Q d v e Psvcholow. 1297-136.

Tnimbo, D., & Miione, F. (1971). Rimuy task paformance as a W o n of

273-279.

Tdv& E. (1983). Elements of qiwdic mnnory. New York: ûxEiord University - Tulvhg E. & Coloth, V. (1970). F n e recaîi oftriliaguPl lists. CoPnitive

Psydmlogxi, 8698.

Tuhring, E., & Thomson, D M (1973). Encoding specincty and d e v a l proasses

in episodic w r y . Psvcho10nicJ Review. 80,3529373.

Tuhrllig E. KipP, S., C e F M , Moswvitch, M., & Hode, S. (1994).

~~c amdbglretrieval Mynmietry in episodic w r y : Positron M o n fiidinp.

Ploacénie, of the National A a u h y of Sience of USA 9 t 20162020.

Wagier, AD., Poidrack, RA, Eldridge, L.L., Desinonci, J E , Hova, GE, &

Gabrieii, JDE (1998). Materiat-spdic lateraüzation ofprehstal .ctiVaton d u h g

episodic enadhg aad retrieval. Neu~~tqprt 9,371 1-3717.

WecWer, D. (1981). Wechsler Ad& Inteili~ence Scaie - R d s e d manual New

Yoric Psych010gicai Corp.

Wechsler, D. (1987). Wedider Merno? Scale - Reviscd manuai . New York:

Psychdogical corp.

Welford, AT. (1 993). The gerontoIogical baiance b e t . In 1. CereiIa, J. Ryb8sb,

idionnation ~rocessinn: Limrts O W. Hoyer, & ML. Commons W.), Nuit œ . dosa. @p. 3-

10). New York Acadernic Press.

Whiting, WL., & Smith, AD. (1997). M e d age-relatecf proceaPiag limitations

in d ad recognition tasks. Py&ology and 1% 216224.

W ' i CD. (1980). Tk~af . r tmt inn i Iresou1:ces , h L S . Nidrrrsan

(Ri.), &enfion and Pdonwice VIl& @p. 239-257). HiUdale, New J- Lawrence

ErlbaumAssociates.

W-, G., & MoscovItch, M (1999). Antemgrade and retroerade d a f k

tesions to f h t a I cortex in rats. J o d o f N " m œ . 19, (2lX 9611-9611.

Zaclcs, RT., & Hasher, L. (1997). Cognitive gaordoiogy ad attentionai inhibition:

A rcply to Burlre rnd McDowd. J o d of G e m ~ o l o ~ P~cbofomccil Scicnc*r 526,

P274P283.

Pm going to turn on the tape, and you wiil hear a f d e voice read 16 différent words, one at a tirne. 1 want you to try to r e m d e r the woids.

Jwt to let you kmw how the tape is ret up: Three bœps wül signai the b m g of the ast and thm beeps wil i also be played at the end of the list. FoUowing this, a nnaber @ce 210), wiii be spolai and 1 want you to stari c o h g backwards out ioud by thnes fiom tht munba. You wiii do this for 15 seconds, untü 1 t d you to stop.

Then, 1 will teil you to s t a r t rralling out loud as many ofthe 16 tuped words as you con You can d the words in any order. You wül have one minute for this r d memory ta& Is evaythiag clear? Do you hve aay questions about what you are suppose to do? 1 wiU ramind you about what p u have to do More erch ta&

Stop countiog bachvards. Begin ncilüng words h m the tape NOW:

Vay good. Now you'ii get a practice session on a sIiehrly cüffèmt task. Ra- üuseeniga~of~rad~dewords.yai'~secalistof1.2and3 syiiable words. For wmple, love, hmmrrr Md bubeque arc wmph of 1.2 rad 3 syilabie words. Ywr task b to bit the dver ky wbeamr a tryIlrMe word is presented on the meen. Alwa)nlrapyourindexfias~rontbs~soyoucan~ywrre~p0~a9soon~yorr notice a word is tsyliabiu. Are you deu on what y w have to do?? ( d e sure they understand whrt syllabIes are). Again we'li be checking your urxlrscy rate. A d , again, Put-way tsrough the esperiment i beep wiû sound A0 More, tbis is just to fnmilianze .. . you with the sound as it wiii be used h the apaimait.l phase.

Baseline distracting task uMAN-MADE/2-svllableH version

Check which baseline distracting task they do first (see order sbeet)

Weil done. We'li now move on to the nexî ta&

Man-made identification as the distractine task

Shoctiy 1 wiU play you a NEW IWt of 16 words tk you h d d try your best to commit to ~IICIIK)~~. As in the pr(Lctiœ session, ya, sbould count baclcwards by tbrees fkom t h e ~ a s h a n d o f t h t w o F d I i s t ,

TEEN, you will go to the cornputer anci start the -MADE test, hining the siiver Lay whenever you notice a man-de word is presented to you. You wüi continue to pafomi the marmade test ALONE, untii you hear the b* sound. This will be your nie to START RECALLJNG as muiy of ths words from the tape that you un rem-, and as More, 1 wül be tape-recordhg yout responacs.

However, tbb is a dual-task eqerhnt, so you MüST CONTINUE DOING THE MAN-KADE TEST WHl[LE SIMULTANEOUSLY TRYING TO RECALL WORDS FROM THE TAPE.

0

1t U VERY important tht you conthme to & the MAN-MADE task WEILE you are trying to r d the words fiom the tape. YOU sbould nlace a 50/50 e m ~ b u b on meh -

Any questions? We wiii now kgia. 1 wiil play the tape, and you sboutd try to commit as meny words to memory 8s you cm.

START a BUZZ CZIE 11111

REMIND Participants to keep dohg ma-made ta& while d g wods

Baseline Recall

This wül bc your nie to START RECALCING as rmuiy of the words b m the tape tht yw caa remember, and as More, 1 wül be tipacecodng your responses.

However, yow ONLY task der the BU22 sound wüi be 10 R E C U L WORDS FRoM THE TAPE That is, the MAN-MADE/2-syiiabfe ta& w9 end, the saeen wiU go biank oace the BU22 sound is hearâ. YOU SHOUID FOCUS ALL OF YOüR ATl'ENïïON ON RECALUNG AS MANY WORDS FROM THE TAPE AS YOU CAN.

Anypuestiot19? Wewülwwbceip IwiUplrythe~ipe,anâyousbouldtryto comxnit as mpay words to rnemory as you can.

Wd doae, ww take a break so you arc +rpdv for mwe w h we get back

Give 4 minute break in lounge

62-svlla ble' identification as the distracting task

However, thU h a dual-taslr so you MUST CONTINUE DOING THE Z-~yIlabk TEST WHILE SIMULTANEOUSLY TRYING TO RECW WORDS FROM TRE TAPE.

It is YERY important that you continue to do the 2-syllible tulr WHILE you are tryine to recaii the words Born the tape. You s h a dace a SU/Sû em~huïs on each la&

Any questions? We wiii now be@ I wül play the tripe, and you shouid try to commit as many words to mawry as you un

REMIND Participants to keep dohg 2-@able task wbüe redhg words

Weil done, now take a break so you arc ready for more w b we get back

GRre 4 minute break in louage

Baselioe distractine task -MAN-MADE OR 2-svllable version

Check which baseline distrachg task they do LAST (see schedule)

GO TO tot# TASK USmG BOTH COMPUTERS.

EXPLAIN TASK ONCE COMPUTERS ARE SET W.

You w i l now be dohg somethiag completdy ciiffietent h m whaî you just finished doing, it invohns üstening to and idaisifymg tonas tbrt you bar CO* h m the cornputer m m *

GREAT

Now I'U h v e you do the tone test d wet again, but tbis time, &a about 15 seconds of dohg the tone task aione, I'll &art the program for tbe MAN-MADE//%SYLLABLE TEST (add subject W: rna.n.de then 2-syl(obie; w u subject #: 2-syllabe thai man- made).

As beforr, You'U have to y v T E S 9 out loud whenever you see a IMAN-MADWf2- SYLLABLE word presented to you, anci 1 wiii Iiit the silver key for you over here on this keyboard.

SO, you must continue to do the tone ta& whüe simuftaneously identiQing MAN- MADEWSYLLABLE words fiom the saeai. Are you dear on what you have to do??

You're gohg to be doing a duai-ta& ~ w h i c h m t ~ ~ t h r t y o u w i l l b e doing two diffkrent tsPLs at the same timt. Pii show you the two Usb separately fint, then we'il combine them. We9B start rrith tk ntl8: mœmry mk.

I'm gohg to tum on the tape, and you wül heat a fimie voice reaâ 16 different words, one at a time. 1 want you to try to remember the words.

Just to la you know how the tape is set up: Thrrebecpswüls i~thckgianineoftheartd~beepswülalsobapLyad.t~ end of the 1Wt FoUowing tbis, a iiumba @ce 210). wül be spoken and 1 want you to start d g ôackwarda out loud by th- ftom that munba. You wül do this for 15 seconds, UntilI tdl you to stop.

Then, Iwiiitdlyou t o r t u t ~ g o u t l o u d a s m a n y o f t h e 16 tqedwordses you aa Yw can c e c d the wor& in MY o r d a . You will have one minute for this lCQll memry tiaL; b avaything d d Do you h v e any quedons about wht you are suppose todo? I d r e n i i a d y o u l b o u t ~ y o u h v e t o d o b e f o ~ ~ h c n s l c

Stop couatiag backwards. Begin recaliing words h m the tape NOW:

Vgy good Now you'll get a @ce SeSgion on a sliebtiy ciiffirent tLpk Rathet th.n seehgastrhgofpicûms, you'ilseealistof 1 , Z d 3 syiiaôlepronowrctibleno* words. Tht is, the letter strings you'lI rea on the icrran do mt mûce up an actuaf worâ, but you WU be abk to sound h ait to y o d s i n c e the nan-wotds in d e to look anci sound like r d words.

For exampie, bove hammet jubeque

ate examples of 1.2, aad 3 Synale non-words. Yom task is to bit the green key wherievcr a trJiiabk non-word is prrseated on the smen Always have y o w hger on t h e g r e e n ~ s o y o u a n d e ~ ~ t t s o c w d e a d e i f i t b a 2 - q l l a b l e 1 1 0 1 1 - w ~ r d . Tbe cornputer won't wait fbr yau to mJu a raponse. so bm your finger &y to prem the key at ail times.

And, again, Part-way through the crpuimtnt r beep wiü round. As beforr, this isjusttofamilianze F I - you with the somd as it wiii k used in the a<paimentd phase.

Baseline distractine task PICTURE animacv/2=svllable~ version

Check which baseline distracthg task they do first (see sehedule)

The aq#rimcnut phse w i l now begk I want yor to know tbat for each session, brand n c r s e t s . t w . r d r w ü i k d Now,Ianmttonndouthowu#inteyourrrrtthe PICTüRE81Û~~~~î2-syuirnetuL W e w a c u l l e c t d d . ~ p n C t i c e d o a s o i a c e w e uaQratmdyou'rejustgattingiiscdtothetasks.F-?i.siï;dlvyour~istbasrm~wbaî

Shortly Iwülplay you aNEWüstof 16wordsthstyoushddtryyourbest to codtomemory. hUithepntCtiœdon,yousbaiMoouptbackwardsbythneshm the rnimber at the end of the wod Est.

~ , y o u w ü l g o t o t h e ~ ~ m p r t e r d r t ~ t t & P I C T U R E i n l l m c y t ~ ~ the green key whenever you notice a pictura tht npnsecitr a nm-made item. You wül continue to pafonn the picture test ALONE, untii y w hear the buzzing soun6

This will be your cue to START RECALLING as m ~ y of the words h m the tape that you an rrmemba, and as More, 1 wüi ôe tapencording your responses.

However, this is a duai-trslr acpaimnt, so you MUST CONTINUE DûING THE PIC'ïURE a n h ~ ~ TEST WHILE SIMULTANEOUSLY TRYING TO RECW WORDS FROM THE TAPE,

It is VERY important that you continue to do the PICTURE tPsk WHLE you are trying to r d the words h m tbe tape. Yau sh p d d ~ h c e a NI50 emnbasis ai, cich

REMIND Participants to keep dohg pichirr tark whüe r e m h g words

GNe 4 minute break in louage

Baselioe Recall

TEEN, you d go to the comrputa and &art tbe PICïüRE raimsCy/2-syWIe test. bithg the grcm ky wimcvct you mtiw a picture of a manmde item D-sSld,Ie wb

word b pmmted to you You wül continue to paform the test &ONE, umü you hear the -- ThU wül be your cue to START RECALUNG as many of the ~or& Born the tape

th.t you can remem&c, and as More, 1 wiîi k tape-recordtig your respollses.

However, your ONLY ta& a f k the BUZZ #Md will be to RECALL WORDS FROM THE TAPE. Thrrt is, the PICTüRE immiCy/2-sylLb1e non-word task wül end, the sueenwülgobiankoncetheBUZZ~isheard. YOUSHOULDFOCUS W O F YOUR ATTENTION ONRECALLJNG AS MANY WORDS FROM THE TAPE AS YOU CAN.

Any questions? We wiii now begh. L will piay the tape, and you should try to commit as many words to mernory as you am.

Weil done, now take a break so you are rcady for mon when we get back.

Give 4 minute break m louage

62-svllable non-word9 identification as the distracting task

N a coma the dhtask part of the upaiient. PU kt yau listen to a new set ofwords onthetapc,thcntatyanm~mdsyEwtha~. BUT,youwa(biwtotrytorccailthe wordr h m the tape wbik SIMULTANEO€JSLY doing the 2-syUaIe non-word wk on tbe ~ ~ l l l ~ ~ f e t ~ PLL READ YOU AN OVEIWEW OF WaAT YOU HAVE TO DO, SO TEiKT YOU ELATE A GENERAL IDEA OF WI3A'S GOING TO EEAPPEN.

Howcver, thW W a dd-task e q m b n î , so you MUST CONTLNUE DOING THE 2 - w TEST WHILE SIhdULTANEOUSLY TRYING TO RECAU WORDS FROM THE TAPE.

It is VERY important tht you continue to do the 2-syIlab1e task WHEE you are tryhg to r d the words f?om the tape. You should niace a 50/50 tm~hmb on each

Any questions? We will now begh 1 wül pliy the tape, and you should try to corninit u muiy words to memory as you can.

START rt B U , CIIE 11111

REMIND Participants to keep dohg 2-syaable task while recaIiing words

Wdl done, now take a break so you are ma& for mon when we get back

Baseline distracting task -PICTURE animacv OR 2-svllable non-word version

Check which baseline distracting task they do LAST (sec schedule)

Ignore the BUZZ TOM tht wiü be bani part-way thrwgh.

ï'mgoing to tumon thetape, mdyaiwülhearafèmaIevoiceread 16 diffaait words, one at a tirne. 1 want you to try to rrmemba the worda

Jui to let you hiow how the tape is set up: Tbree beeps will signai the beguming ofthe îist rad tbne beeps wül aiso be played r the end of the W. Foiiowùig this, a aumba @ce 210). wüt be spoken Md 1 wuit you to start comhg b r d w d s out ioud by Uuees fiom tbat number. You win do b i s for 15 seconds, umil I t d you to stop.

~ I w i i l t d y o u tostartredhgout toudrcimrnyofthe 16tspedwordsas you cari. You CM 4 the words in my ordtr. You win bave one alnute for this r d memory tadq Ia tvaything CM Do y w have my questions about what you an suppose to do? 1 will remiind y w abut wbet you have to do Mre each WC.

Stop countiag backwardss Begin recalling words fiom the tape NOW:

For example, bove hammet jubeqw

ut eacamples of 1,2, Piad 3 syUable non-words. Your uslr is to hit the green k y whaever 8 2-ayilrbie non-nord is presented on tbe scran Aiways have your hger on the green k y so you a n make your response as sooa decide ifit W a 2-syîiable non-word. The cornputer won't wait for you to make a response, so h v e your finger ready to press thekeyatrlttimes.

Baseline distractine task PICTURE SUE/2-svllablew version

Check which brrsethre distracthg task they do fmt (see schedule)

Well dune. We'il mw move on to the next task

This will k y w r aie to START RECALLING as msny of the words fkom the tape tbet y w crn temember, and as More, 1 wül k taprecordîng your nsponscs.

However, this is a d d t a a k atpaimnt, so you MüST CONTINUE DOING THE PICTURE SIZE TEST WHILE SlMULTANEOUSLY TRYING TO RECALL WORDS FROM TEIE TAPE.

It is VERY hiportant that you w h e to do the PICTURE SIZE ta& WH[LE you are tryhg to the words fiom the tape. Yoa show nhce 8 a/- tm~hasir on sasumk

REMIND Participants to keep dohg p i ~ a a e 9ze tuk whüe recalfing words

Baseline Recall

THW, you wül p to the amputer rad start the PICîüRE SW2-@able test, hitt@ the green key wknever you notice a pictura b i w thrn a mo~or/2-syiIabIe non- word is p h to you. You win continue to pafom t k test ALONE, until yw hear the buzPqe-

This wiU be your cue to START RECLUJ*ING as m~iiy of the worb f?om the tape that p u cm cemembef, and as kforc, 1 Win be taperecordhg your responses.

Howmr, your ONLY trrk a f k the BUZZ round wül be to RECALL WORDS FRoM THE TAPE. Thrt is, the PICTURE SIP3I2-syllible mwword ta& will end, the screenwülpbiankoncetheBUZZdbhearci. YOUSHOULDFOCUSAUOF YOUR ATTENTION ON RECALLING AS MANY WORDS FROM THE TAPE AS YOU CAN.

h y questions? We wül now be@a I wiil play the tape, and you shouid tty to commit as many words to memory as you can.

62svllable non-word9 identification as the distracting task

However, bis tbio a dual-taslr a<paimaat, so you MUST CONTINUE DONG TEIE Zrytkbk TEST WEKE SIMUtTANEOUSLY T R W TO RECALL WORDS FROM TfIE TAPE.

1 b VERY inportut t h you continue to do the 2-syUable task WHILE you are trying to r d the words h m the tape. You should ahce 8 SOMO em~huis on ucb la&

Any questions? We wül now bcgin. 1 wül play the tape, and you shouid try to cornmit ~manywordstomemoryrryouaul

REMIND Participants to keep doing 2-sylLbte task wMe recaUîng words

Give 4 *te break in lounge

Baseline distracting task-PICïURE SIZE OR 2-svllable non-word version

Check which baseline distrading trisk they do LAST (see schedule)

Ignore the BUZZ TONE that will be h d part-way tbrough.

Weii done.

You'n gohg to be dohg a wbich means that you wül be doiugtwo Wérenttuksattbesimttimc. P ~ r b o w y a , t b e t w o t a s k s ~ n n f then m'Il combine thea. We'l rtut wiU the & memoay ta&.

~mgo~totumonthetrpe.indyouwiuharanrmkvoicenad l6diffaent worQ, one at a tmie, very dowiy. 1 want yw to try to rememk these words.

Jiut to let you h o w how the tapa is set up: Three beeps wül signai the keinaing of the ht and ttves beeps wül also be ployed at the end of the list. Fonowing this, a numba (üke 210). wül be spokm and 1 want you to stlvt d g b.ckwuds out bud by threes h m that number. You wiii do this for 15 seconds, umil I teil you to stop.

~IwinMyoutostvtnedliigoutIoudumrnyofthe 16tapeàwordsu p u aa You can recaii the words in u y order. You wül have one minute for this mal ==Y*

1s evaything clear'? 1 wiU remid you about wbat you have to do More each task.

Record recd O- casseate

Vay good Now you'll get a @ce session oa a di- diffefeaf tasic. RitLer than saeing a ibt ofanimal and mzlJEmade worQ. youll see a lisr of 2-digit itumbas. Some ofthcaumbsnwillbea~q minû, 2,4,6,or8.admmewilt beodd, endiagm 1,3, 5,7, or 9.

Your tsslr is to hit the gnan key whenever an odd-digît is presentd on the saeea. Aiwa. keep your index finger on the k y so you CM mh your rasponre as mon as you notice an odbdigit

A& we'n be checking your icnmcy rate. And, @II, Put-way through the wtpcrimeat a bœp wüi round. As WÔre, thW isjwt to hmilune

* * . you wiîh the sound as

it wiii be uced in the exparnientai phase.

Instnictions for Baseline (sinele-task) iiMAN-MADE (or odd-dieitY' measure Check which single-task baseline they do first (see order sheet)

Wdl done. We'H aow move on to the Uext îaak

Man-made identification as the distractine task

BUT,youwillhvetotiyto~tbeworbâomthe1.pewbüeSIMULTANEOUSLY dohg the rmwiide tasic on the cornpiter. IU READ YOU AN OVERVIEW OF WHAT YOU HAVE TO Dû, SO TRAT YOU HAVE A GENERAL DEA OF WHAT'S GOING TO HAPPEN.

Shore 1 wül piay you a NEW oit of 16 wods thrt y w sbwld try your best to commit to mcmory. As in the practice sessi04 you shouid count bachards by threes from themjmberatthedofthewordüst.

TEEN, you wili go to the coinputer and start the MAN-MADE test, hitting the green key whenewr y w notice 8 man-made w o d h presented to you. You d continue to pedonn the mumade test ALONE, untii you hem the btdng souad. This wüt k your cue to STARI' REUIUNG as many of the words âom the tape that you can mnember, and as More, 1 d be trpoc8cording your respo~lses.

HO-, îhis is a ~~ apzhmt, so you MüST COBlTWüE DOMG TXE MAN-MADE TEST WHILE SMULTANEOUSLY TRYING TO RECALL WORDS FROM THE TAPE.

It is VERY important th.t y w continue to do the MAN-MADE task WHILE you are üyiag to d the words âom the tape. yoa should ~ I a c e a 50150 e m ~ h u i s Qn each fuk

Any qwstions? P l a y W w a d r n

Full Attention durioe recall

AGAIN, IwülplayyouaNEWlirtof 16 wordr th tyou~dtryyourks t to commit to manory. As in the pfadice d o n , you sbould unrnt backwarâs by threes fiom thenumberattheendofthewordiia

T H M , y o u f i g o tothecompitariad startt&(seeordersheet: ifv- task bae y '-mades; ifvehu, tdc hem is 'odd-dr*gitP) MAwMADE (or odddigit) test, hitthg the grrai key wheneva y w notice a 'm~i~~-madc' word (or oâd-dîgit) LI presented to y o u You will continue to pafonn the test ALONE, Uatü you hcar the b&iz sauné

This wiîl be your me to START RECALLING as many of the wods h m the tape that you cm rememks and as More, 1 wül be tape-recorcüng your respollses.

However, yyour ONLY task &a the buzz souad wül be to RECW WORDS FROM THE TAPE. That is, the MAN-MADE (or odd-digit) ta& wili enâ.

The screen wüi p blpnlr once the BUZZ s d is W. YOU SHOWLD FOCUS ALI, OF YOUR ATTENTION ON RECALLING AS MANY WORDS FROM THE TAPE AS YOU CAN.

Any questions? We wül now begh 1 Win piay the tape, and you shouid try to commit esm~nywordstomamoryasyouc9n

'odddieit9 identification as the distractine task

BUT, you win bave to try to r c d the wods h m the tape whik SLMULTANEOWSLY dokg tba &-digit tult on the compba. IZL READ YOU AN OVERVEW OF WEUl YOU HAVE TO DO, SO THAT YOU HAVE A GEBERAL DEA OF W a A T S GOING TO HAPPEN.

This wül be your aie to START RECAIUNG u mrny ofthe worb h m the tupe thtyou~rsmmkr,iad~befO~Iwülbetrpocec~r~yourrrspoases.

Howcver, tbU h a dual& eilpaima9 ao you EIIUST CONTINUE DOING THE odddigit test WHILE SIMüLTANEOUSLY TRYLNG TO RECALL WORDS FROM THE TAPE.

It U WRY importuit that y w continue to do the ociddigit ta& WHILE you are trying to r d t&e words fiom the tape. You shouid dace a !!O/SO e m ~ b u i r on mch a?&

Any questiom? Physl3Ui)wardEst

Now that you don't have to do two things rt the slme time, am you d any mon words Born the study list?

Weil done, now take a b n i k so you are ready f9r more when we get badc Give 5 minute break

Baseline (single-task) <CMAN-MADE (or odd-di@tln measnre Check which distracting CnsL they do (see order sheet)

Wd donc.

hutmctiou for auditory CRT tuk perlormed coacurrtatiy with eaeh dirtircfjng tuk

Your 3 finga h u l d always k on the k q s so bat you aan make your nsponse as qyiciciy andasarr?l~rntelyrrpomible. Assooa~yaubitaLytoidmtit;/~theneidtonewin wme on, and you mua respond ri@ awry. As such, this U a CONTINUOUS trdE since you'll ahivays be identifjing a tone.

The tuk lasts for 2 minutes ( l S actually) so txy to stay f i d .

GREAT

Now riî hnn you do the tom test in over again, but this the, after about 15 seconds of doing the tom task done, I'U start tbe program fm the WordaaimrCy 1 or word-2-syiIabIe / or PICTURE Mimecy/ or pi- S W or ?-SYLLABE non-word/ or odd-digit id TEST.

You'll have to gav 'YES' out loud whenewr you see a word denotiag a mpn-made item/ or 2-syüable word I or pi- of an item that is rnPPmadd or PICTURE of an item tna ik E@ger tnpn an average compuaer ~ ~ ) n i f o r h r ZlSYLtABLE ma-worU or add-dgît ispresaitedto yo~oadIwülhitthegraakyforywo~hanonthWkeybwd.

SO, younnutcontinueto do thttonetrdtwhk simitrneouslydoingibe &ra&gtasLonthesaeen Areywdarroawhtyaihvetodo??

Appcndû 2.1

Wpk data cdltction ahœt for fm reaü

Orderof Sessions:

Appcidu 3.1

S.mple stimuli for dirtrrcting tub in Expuiiiieat 1

feather

Uable nonsense word tapk

-= Siapk 3timPü for dirardng tub in Experiaienb 4,s and 6

house